Maintenance Pack NB_51_4_M.sgi.tar provides fixes for VERITAS NetBackup (tm) Enterprise Server 5.1 on SGI IRIX servers. NetBackup UNIX Add-on products, Database Agents and Java GUI have separate Maintenance Packs.
| Article:TECH45020 | | | Created: 2005-01-22 | | | Updated: 2006-01-01 | | | Article URL http://www.symantec.com/docs/TECH45020 |
Problem
Maintenance Pack NB_51_4_M.sgi.tar provides fixes for VERITAS NetBackup (tm) Enterprise Server 5.1 on SGI IRIX servers. NetBackup UNIX Add-on products, Database Agents and Java GUI have separate Maintenance Packs.
Solution
NB 5.1GA Pack NB_51_4_M
README
November 21, 2005
Requirement: NB_CLT_51_4_M
================================================================================
This Maintenance Pack provides fixes for VERITAS NetBackup (tm) UNIX servers.
NetBackup UNIX Add-on products, Database Agents and Java GUI have separate
Maintenance Packs.
Symantec recommends that the backing up of active file systems be avoided,
or the use of snapshot technologies be implemented. The directory structure
reported back from the file system (to NetBackup) may not contain all of the
files available during the time of backup. NetBackup will not report errors
in many cases where the file's existence is not known to NetBackup as reported
by the file system.
================================================================================
=================
PACK DEPENDENCIES
=================
-- NB_CLT_51_4_M_<6 digit number>.tar must be installed before this
Maintenance Pack is installed.
-- Installation of this Maintenance Pack requires version 1.19.4.23 of
the Vrts_pack.install script.
I. DOWNLOAD INSTRUCTIONS
II. KNOWN ISSUES
III. INSTALLATION INSTRUCTIONS
IV. UNINSTALL INSTRUCTIONS
V. DESCRIPTION OF PROBLEMS FIXED
Current Pack
NB_51_3_M
NB_51_2_M
NB_51_1_M
=========================
I. DOWNLOAD INSTRUCTIONS
=========================
1) Download the NB_CLT_51_4_M_<6 digit number>.tar and
NB_51_4_M_<6 digit number>.<server>.tar files into the
/tmp directory,
where <6 digit number> is an internal tracking identifier
where <server> is alpha_5, hp_ux, linux, rs6000, sgi, solaris
NOTE: NB_CLT_51_4_M_<6 digit number>.tar has the client binaries and
NB_51_4_M_<6 digit number>.<server>.tar has the server binaries and BOTH
must be installed.
2) Extract the NB_CLT_51_4_M_<6 digit number>.tar and the
NB_51_4_M_<6 digit number>.<server>.tar files.
tar xvf NB_CLT_51_4_M_<6 digit number>.tar
tar xvf NB_51_4_M_<6 digit number>.<server>.tar
NB_51_4_M will create the files:
VrtsNB_51_4_M.README
VrtsNB_51_4_M.<server>.tar.Z
VrtsNB_51_4_M.postuninstall
VrtsNB_51_4_M.postinstall
VrtsNB_51_4_M.preinstall
Vrts_pack.install
NB_CLT_51_4_M will create the files:
VrtsNB_CLT_51_4_M.README
VrtsNB_CLT_51_4_M.tar.Z
VrtsNB_CLT_51_4_M.postuninstall
VrtsNB_CLT_51_4_M.postinstall
VrtsNB_CLT_51_4_M.preinstall
==================
II. KNOWN ISSUES
==================
Description:
Solaris 10 is now supported as follows:
Hardware Support
---------------------------------------------
SPARC NetBackup Server and Client
Intel x86 NetBackup Client
The above support is for base OS support only. With NetBackup 5.1 MP2,
the "Zones" or "Containers" feature introduced in Solaris 10 is not
supported. NetBackup is supported on a system pruning global zone
only, any additional local zones or containers are not supported at
this time.
The support of NetBackup Client for Solaris 10 x86 on AMD Opteron
servers is not supported with NetBackup 5.1 MP2, but the support may
be announced at a later time based on successful test results.
Use of NetBackup Advanced Client methods are supported on Solaris 10
beginning with VERITAS Storage Foundation Suite version 4.1 releasing
in 2005. Check the Sun/VERITAS web-site for availability.
Solaris 10 is not supported on the base CD-ROM version of NetBackup 5.1.
There are known connection and Java GUI issues that will be encountered
if you attempt to run the NetBackup 5.1 GA version on Solaris 10. This
is due to a new inetd design method introduced in Solaris 10.
Therefore, this NetBackup 5.1 MP2 patch update must be applied.
The following script must be run after this patch is installed in order
for "Solaris Solaris 10" or "Solaris Solaris_x86_10" to show as a client
selection in the drop down list for backup policies:
/usr/openv/netbackup/bin/goodies/new_clients
Upon running this script, Solaris 10 choices will be available in the
drop-down menu for Solaris 10 clients.
===============================
III. INSTALLATION INSTRUCTIONS
===============================
For Maintenance Pack installation on a UNIX Cluster Environment:
NOTE: Click on the "Download Now" link, near the bottom of this document
prior to running the following installation procedure for this pack.
1) Ensure that prior to installing the maintenance pack, NetBackup is at
release level 5.1 and configured to run in a cluster.
2) Freeze the NetBackup group (This will avoid a 'failover' during a patch
installation).
3) Install this Maintenance Pack on the inactive node(s) of the cluster
(follow steps 1-3 below).
4) Install this Maintenance Pack on the active node of the cluster (follow
steps 1-3 below).
5) Unfreeze the NetBackup group.
--------------------------------------------------------------------------------
There are two ways to install the client maintenance pack software.
1. Remote Installation: Loads the software on a master server with the intent
of pushing client software out to affected clients.
2. Local Installation: Loads and installs the software only to this local
machine.
Remote client install:
As root on the NetBackup Master Server:
1) This pack contains a full release of the IBMzSeriesLinux 2.4.21 client. To
install, you must first create the following directory:
mkdir /usr/openv/netbackup/client/Linux/IBMzSeriesLinux2.4.21
If you intend to install the Encryption libraries for this client type,
you must create the following directory:
mkdir /usr/openv/netbackup/crypt/Linux/IBMzSeriesLinux2.4.21
2) Install NB_51_4_M and NB_CLT_51_4_M Maintenance Pack binaries.
cd /tmp
/bin/sh Vrts_pack.install
3) Restart daemons.
/usr/openv/netbackup/bin/initbprd
/usr/openv/volmgr/bin/ltid -v
4) Update the NetBackup clients with the update_clients script.
/usr/openv/netbackup/bin/update_clients <hardware> <os>
where <hardware> <os> is one of the following:
ALPHA OSF1_V5
HP9000-700 HP-UX11.00
HP9000-800 HP-UX11.00
HP9000-700 HP-UX11.11
HP9000-800 HP-UX11.11
Linux IBMzSeriesLinux2.4
Linux IBMzSeriesLinux2.4.21
INTEL FreeBSD4.5
Linux RedHat2.4
MACINTOSH MacOSXS10.2
RS6000 AIX4.3.3
RS6000 AIX5
SCO UnixWare7.1
SGI IRIX65
Solaris Solaris7
Solaris Solaris8
Solaris Solaris9
Solaris Solaris10 *
Solaris Solaris_x86_7
Solaris Solaris_x86_8
Solaris Solaris_x86_9
Solaris Solaris_x86_10 *
* - For "Solaris Solaris10" or "Solaris Solaris_x86_10" to show when
configuring clients into a backup policy, the following must be
executed on the server:
/usr/openv/netbackup/bin/goodies/new_clients
Note: The /usr/openv/netbackup/bin/update_clients command without
any parameters will update all the UNIX clients.
Note: When updating an RS6000 client, there may be circumstances
where update_clients will fail with an error similar to
this:
Couldn't open /usr/openv/lib/libVmangle.so on client
Client open errno = 26
If this happens, execute /usr/sbin/slibclean on the client
to be updated and re-run update_clients.
If the client (CLT) .Z file and README exist in the installation
directory during the installation of the server maintenance pack,
the Vrts_pack.install script will install the client maintenance
pack automatically. The client maintenance pack will NOT be
installed automatically during a reinstall of the server
maintenance pack.
Additional Notes:
If non-root administrators use the GUI only, the nonroot_admin
script no longer needs to be run. If the non-root administrators
use the command line or bpadm, the group and file permissions
will have to be changed manually on the NetBackup binaries.
Users can write their own script. The script is being phased
out because there is a slight security risk that non-root users
may be able to execute NetBackup commands only because those users
are part of a group that is allowed to execute NetBackup commands.
For "Solaris Solaris10" to show when configuring clients into a backup
policy, the following command must be executed:
/usr/openv/netbackup/bin/goodies/new_clients
Local client install:
The install script will determine if a local client install is appropriate
and choose the appropriate client type to install.
As root on the NetBackup client:
1) Install NB_CLT_51_4_M Maintenance Pack binaries.
cd /tmp
/bin/sh Vrts_pack.install
NOTE: It is not possible to install the new Linux client
(LinuxIBMzSeriesLinux 2.4.21) locally. It must be installed on the
Master server and pushed to the client.
===========================
IV. UNINSTALL INSTRUCTIONS
===========================
Note: This will ONLY uninstall the Maintenance Pack from your local machine.
1) Close the NetBackup user interfaces.
Make sure the NetBackup server has no active jobs running (for
example, backups, restores, or duplications).
If a database agent is being used, such as Oracle,
ensure that the database services are stopped.
2) Change directory to the patch save directory.
Substitute the pack name for ${PACK} in the following command:
cd /usr/openv/pack/${PACK}/save
3) Run the un-install script:
./Vrts_pack.uninstall
4) Verify that the pack uninstalled successfully by checking
/usr/openv/pack/pack.history.
5) If update_clients was run after the pack was originally INSTALLED,
run it again after that pack is successfully UNINSTALLED.
6) If necessary, restart the NetBackup and Media Manager daemons:
/usr/openv/netbackup/bin/goodies/netbackup start
=================================
V. DESCRIPTION OF PROBLEMS FIXED
=================================
The following are descriptions of the problems fixed.
Please read the entire document before installing.
README Conventions:
Description
Describes a particular problem contained in this Maintenance Pack.
** Description **
Describes a problem that can lead to potential data loss. Please
read these problem descriptions carefully.
Workaround
Any available workarounds to a problem are also listed. Workarounds
can be used INSTEAD of applying the patch, however, Symantec strongly
recommends the "best practice" of being at the latest patch level.
Additional Notes
Any additional information regarding this problem or feature is included.
=============
Current pack
=============
================================================================================
Etrack Incident = ET342058
Description:
If a machine has NetBackup and a Maintenance Pack (MP) installed, and an
uninstall is performed of the MP that requires a reboot, and NetBackup is
uninstalled before the reboot, the DLL named msvcp70.dll is removed. The
installer has been modified to ignore MFC DLL in the Windows system32
folder.
Workaround:
To avoid this issue, perform a reboot when you are requested after first
uninstalling the MP and before you uninstall NetBackup.
--------------------------------------------------------------------------------
Etrack Incident = ET263530
Description:
For Solaris10, a change needed to be made to make sure that /etc/services
was updated before inetd.conf. This was only an issue on UNIX servers and
only if YP was not configured.
--------------------------------------------------------------------------------
Etrack Incident = ET312135
Description:
NetBackup 5.1 clusters need to be aware of NetBackup Operations Manager.
Without this patch NetBackup Operations Manager will not work in a
NetBackup 5.1 cluster.
Workaround:
The user could add the following lines to
/usr/openv/netbackup/cluster<cluster_platform>/<platform>_NBU_RSP.
DIR=var mkdir
DIR=var/global mv
LINK=var/global
--------------------------------------------------------------------------------
Etrack Incident = ET374817
Description:
A user currently cannot enter more than 99 dates into a calendar
schedule. To resolve this issue, increase the size of some of the arrays.
--------------------------------------------------------------------------------
Etrack Incident = ET354621
Description:
DSSU relocation would fail after upgrading to NetBackup 50_MP5 or 51_MP3.
--------------------------------------------------------------------------------
Etrack Incident = ET341834
Description:
Disk Staging cleanup was not occurring, nor was it notifying the user
quickly enough if the Disk Staging Unit was a NAS mounted file system.
--------------------------------------------------------------------------------
Etrack Incident = ET346373 ET366137 ET322959
Description:
Corrected the display of new ACS vendor media types T10000T1, T10000TS,
and T10000CT. In addition, corrected the allowable types that ACS vendor
media type LTO_400G.
--------------------------------------------------------------------------------
Etrack Incident = ET361923
Description:
Some character sets do not support upper case even though the underlying
language does. This caused some wildcard constructs to fail because the
upper and lower cases were not distinguishable.
Workaround:
To resolve this issue, the wildcard constructs were reworked so [] and '-'
could not be used with upper and lower case.
--------------------------------------------------------------------------------
Etrack Incident = ET325076
Description:
Backups of NetLink 2.x and higher files failed.
--------------------------------------------------------------------------------
Etrack Incident = ET311173 ET223766
Description:
If the add_slave_on_clients command is unable to add the media server
entries to one client, it may incorrectly update the media servers on
subsequent clients. Typically, only one media server will be added to
the remaining clients.
Workaround:
To avoid this issue, perform one of the following two work-arounds:
- Use the Host Properties interface to update the server list for
the clients.
- Use the bpsetconfig admincmd to update the server list for the clients.
--------------------------------------------------------------------------------
Etrack Incident = ET364429
Description:
** A stack trace showed that bpbrm was in the exit() system call, causing
bpbrm to dump core. This does not affect backups or restores.
--------------------------------------------------------------------------------
Etrack Incident = ET336219
Description:
The removal of a copy of a backup image was not occurring when the copy
expired. This occurred when the operating system was Linux and the file
system (where the NetBackup database resided) was VxFS.
--------------------------------------------------------------------------------
Etrack Incident = ET355614
Description:
If a backup is being run on an NFS file system and the NFS server is down,
the backup would hang.
--------------------------------------------------------------------------------
Etrack Incident = ET342043 ET207782
Description:
Restoring a large number of Windows files to UNIX would lead to being
overwhelmed with .SeCuRiTy files. (These files are from the Windows
security data that is invalid for UNIX.)
In addition, a Windows to UNIX restore did not indicate success because
of a failure to restore the Windows security data to UNIX.
Workaround:
To avoid these issues, do not restore Windows files to UNIX, and if you do
attempt to do this, make sure you clean up the .SeCuRiTy* files.
--------------------------------------------------------------------------------
Etrack Incident = ET352808 ET343561
Description:
Reverse host name lookup is required for outgoing NetBackup connections.
With this patch, outgoing connections from NetBackup servers that are not
configured for NBAC will not require reserve host name lookup.
--------------------------------------------------------------------------------
Etrack Incident = ET366163 ET355610
Description:
In certain firewall configurations, multistream backups from a Linux
client may fail with a status 25 if the vnetd connect-back is used.
--------------------------------------------------------------------------------
Etrack Incident = ET326978
Description:
After getting a, "client hostname could not be found" error, communication
between bpbrm and bpsched would fail. The bpbrm log contained the following
messages:
18:08:51.752 [24593] <2> put_strlen_str: cannot write data to network: Bad
file number (9)
18:08:51.752 [24593] <16> bpbrm send_parent_msg: could not write
COMPLETE_MOUNT V02107 0 message to stderr
--------------------------------------------------------------------------------
Etrack Incident = ET316059
Description:
Resumed restore from a number of images that involved a number of tapes
fails with the following error messages in the bprd log:
01:27:54.837 [557110] <2> put_length_bytes_nonblocking: cannot write data
to network: There is no process to read data written to a pipe. (32)
01:27:54.837 [557110] <2> write_files_to_active_sockets: Could not write
to file list socket 9 - connection dropped or not connected. Errno = 32:
There is no process to read data written to a pipe.
--------------------------------------------------------------------------------
Etrack Incident = ET341332
Description:
If using Multiple Data Streams, there would possibly be a long delay in
scheduling jobs. The bpsched log would show many hosts_equal calls that
were contributing to the delay.
--------------------------------------------------------------------------------
Etrack Incident = ET330685
Description:
Canceling a queued Vault job caused all backups to cancel.
--------------------------------------------------------------------------------
Etrack Incident = ET321736
Description:
The use of a closed socket error message was filling the scheduler log.
--------------------------------------------------------------------------------
Etrack Incident = ET322851 ET322817
Description:
When a server includes multiple network interfaces and is configured for
NBAC, other NetBackup machines may reject connections from the server
with authorization failures. This patch insures that the server presents
the correct credential to the other NetBackup machines.
--------------------------------------------------------------------------------
Etrack Incident = ET336943
Description:
A user backup could not be cancelled using the Activity Monitor.
--------------------------------------------------------------------------------
Etrack Incident = ET347554
Description:
On Windows platforms only: The scheduler would enter an infinite loop after
A read message queue failure occurred.
--------------------------------------------------------------------------------
Etrack Incident = ET349783
Description:
The scheduler would enter an infinite loop if all of the following
occurred:
- The backup is checkpoint enabled.
- The backup is twinning.
- One of the twined copies failed.
- The non-failed copy exits with a failure status.
--------------------------------------------------------------------------------
Etrack Incident = ET310386
Description:
Previously, there was a logic error that prevented customers from using
cluster_config -r to add robots for monitoring. This patch fixes the error.
Workaround:
The avoid this issue, add the name of the robotic daemon directly to the
PROBE_PROCS line in the <platform> _RSP file.
--------------------------------------------------------------------------------
Etrack Incident = ET363305
Description:
Adding a job to a multiplex group failed to check if a HOLD_134 exists on
the multiplex storage unit. This resulted in excessive 134 status backups.
--------------------------------------------------------------------------------
Etrack Incident = ET366178
Description:
Field 11, of the 'bpdbjobs -all' output, did not reflect the correct end
time of the last try.
--------------------------------------------------------------------------------
Etrack Incident = ET314632
Description:
A change has been made so the scheduler will no longer dump core.
--------------------------------------------------------------------------------
Etrack Incident = ET325883
Description:
The Activity Monitor would continuously display 'positioning' for
multi-fragment backups.
--------------------------------------------------------------------------------
Etrack Incident = ET316646
Description:
The scheduler performance suffered if the media server connection is
intermittent. The problem is aggravated when schedule continuously polls
the media server for NetBackup version information.
--------------------------------------------------------------------------------
Etrack Incident = ET352026
Description:
If an attempt to create a binary .f file is done with file entries without
user or group data, the attempt would fail but leave a zero-length .f
file. This caused a subsequent attempt to create the .f file in ASCII
format.
--------------------------------------------------------------------------------
Etrack Incident = ET315664
Description:
If you attempt to perform a cat_convert on an image where the policy name
contains a period (or full-stop) and the letter [f] follows immediately
after the period (or full stop), then the cat_convert says the file does
not exist.
Workaround:
To avoid this issue, do not create policies with names that contain ".f".
--------------------------------------------------------------------------------
Etrack Incident = ET311559 ET316451
Description:
Backup jobs with multiple streams where the individual stream selections
start with the same path can produce image files which, when browsed, do no
list all of the files in the backup.
Also, .f files created in NetBackup 4.5 versions for non-standard,
non-Windows clients may not show all files in the backup when browsed
under NetBackup 5.1.
--------------------------------------------------------------------------------
Etrack Incident = ET315663
Description:
Policy names containing ".f" cause premature EOF on Full TIR Backups on
the first try of jobid and when following an incremental.
Workaround:
To avoid this issue, do not use policy names that contain ".f".
--------------------------------------------------------------------------------
Etrack Incident = ET375771
Description:
If many Flashbackup images are created using an ASCII image format,
cleanup fails with the message, "too many open files".
Workaround:
To avoid this issue, use the binary .f file format to create the
Flashbackup images.
--------------------------------------------------------------------------------
Etrack Incident = ET371664
Description:
The synthetic backup process would begin reading the wrong block number
from tape during the synthetic backup image creation, causing the
synthetic backup to fail. This would happen when there was more than
one tape drive available to read the component images.
The following sample log message is a symptom of this failure:
bptm 12090 first XFERBLOCK indicated block number of 478341120 but initial
positioning was for block 193352448
This message indicates an, “unexpected block number read,” will be seen in
both the job details for the synthetic backup job, and in the bptm log.
Workaround:
To avoid this issue, down drives so that only two drives are available for
creating the synthetic backup - one drive for reading and one drive for
writing.
After the synthetic backup job mounts the first volume to be read and
mounts the volume to be written, you can safely increase the number
of “up” drives.
--------------------------------------------------------------------------------
Etrack Incident = ET380235
Description:
If the NOexpire touch file was set, then the NetBackup database image
cleanup was disabled. However, media would still expire. The media no
longer expires if the NOexpire touch file is set.
Bpexpdate -deassignempty, which deassigns media if there are no
corresponding images in the image database, is disabled.
The NOexpire touch file is set in the following locations:
For UNIX:
/usr/openv/netbackup/bin/NOexpire
For Windows:
VERITAS\NetBackup\Bin\NOexpire
--------------------------------------------------------------------------------
Etrack Incident = ET355414 ET499608
Description:
Synthetic Backup would produce an image that contains only the last
Incremental if a previous full synthetic backup (from incremental backups)
was manually expired.
--------------------------------------------------------------------------------
Etrack Incident = ET314637 ET405575
Description:
"bpdbm -consistency" failed to move out corrupted image files in
"catstore". It also failed to move out the zero byte image files
(which should have been handled by cleanup but it also failed).
Additional Notes:
On UNIX platforms, “bpdbm consistency” failed to move out the corrupted
image files if /usr/openv/netbackup/db and
/usr/openv/netbackup/db/images/<client> were not on the same volume. In
such case, the corrupted image files will be left to their original
locations.
--------------------------------------------------------------------------------
Etrack Incident = ET379404
Description:
The C_GET_HOSTINFO bprd request (from bpgetconfig -g for example) would
fail if there was not a cipher file on the client/host requested.
To avoid this issue, make sure a non-empty string is passed between the
dbmisc APIs, bprd, and bpgetconfig for the cipher list even if encryption
is not configured on the client.
--------------------------------------------------------------------------------
Etrack Incident = ET365011
Description:
When a C_GET_HOSTINFO request is made from a pre-NetBackup 6.0
installation, the [expected] request type and client name are appended to
the MAGIC string. For NetBackup 6.0 and later, these two fields are sent
separately. Infer the level of the requestor by counting the MAGIC string
arguments, and retrieve the type and client name accordingly. Once
inferred, ensure that the 'acks' that are sent to the requestor are in
synch with the requestor’s expectations.
--------------------------------------------------------------------------------
Etrack Incident = ET378815
Description:
The scheduler predict was not functioning correctly. The input date was
decremented so that it never exceeded the current date.
--------------------------------------------------------------------------------
Etrack Incident = ET371647
Description:
An XBSA application would hang during a shared memory backup (such as, the
client and the media server are the same machine) if a problem occurred
that caused the bptm to exit. The BSASendData() call did not return.
--------------------------------------------------------------------------------
Etrack Incident = ET340730
Description:
If the bpgetmedia options -h and -M were both specified, the -h argument
would always inherit the -M argument, and the original -h value would be
lost.
--------------------------------------------------------------------------------
Etrack Incident = ET323264
Description:
bprd could not clean a vault session directory that contains files larger
than 2 Gigabytes.
Additional Notes:
To resolve this issue, use lstat64() for LARGEFILES platforms in the
recursive_dir_delete() function. Used primarily by bprd for periodic
deletion of log files (and vault session files)... and by bpdbm for
removing expired back-ups.
--------------------------------------------------------------------------------
Etrack Incident = ET386569
Description:
Schedules that specify "instant recovery" should not have backups that
multi-stream. It is possible that a policy may try to multi-stream in
the following circumstances:
- Frozen image enabled.
- PFI enabled.
- Multi-streaming enabled.
- Policy contains schedules with instant recovery enabled.
--------------------------------------------------------------------------------
Etrack Incident = ET313352
Description:
Disk staging only submits one duplication job at a time so it was unable
to make use of idle tape drives.
Additional Notes:
Up to four bpduplicate jobs are submitted at a time. The duplication jobs
contain images grouped into jobs of 25 Gigabytes. By creating files named
MAX_STAGING_JOBS and STAGING_JOB_KB_LIMIT in the NetBackup directory on the
master server, these defaults can be adjusted as needed. For the
STAGING_JOB_KB_LIMIT, the value is in kilobytes and the default is 26214400
(25 gigabytes). A smaller value causes more jobs to be submitted, but
there will only be MAX_STAGING_JOBS running at a time.
--------------------------------------------------------------------------------
Etrack Incident = ET374768
Description:
ACS robot inventory may timeout when ACSLS server is overloaded.
--------------------------------------------------------------------------------
Etrack Incident = ET392051
Description:
Checkpoint-enabled user backups were performing an automatic retry
(limited by the failure history). But user backups should not auto retry.
--------------------------------------------------------------------------------
Etrack Incident = ET373913
Description:
The TPC selection process does not allow for selecting parallel mover
devices when available to improve performance and when multiple backups
are active.
Additional Notes:
This path adds a "GROUP" keyword in mover.conf. The GROUP keyword has a
parameter that is the name of a file in /usr/openv/volmgr/database that has
a list of mover devices that have the same capability. Each bptm that
executes the GROUP keyword will select the next device in the file in a
round robin fashion. A separate file is maintained to keep track of the
next device index.
--------------------------------------------------------------------------------
Etrack Incident = ET374770 ET388367
Description:
Tlhcd may dump core when running dump_tables from ltid -tables command.
Tlhcd needs to check for a media ID with all spaces for that ID to be the
last media in the list that is returned from the robot controller.
--------------------------------------------------------------------------------
Etrack Incident = ET394946
Description:
A restore from a FlashBackup image hung on a Linux server with an
indication of a double free of a memory allocation.
--------------------------------------------------------------------------------
Etrack Incident = ET394942
Description:
User backups or archives that do not find an active policy/schedule to use
were marked as resumable, but without the policy/schedule information, the
job could not be resumed.
--------------------------------------------------------------------------------
Etrack Incident = ET394948
Description:
A duplication may have been reported as successful if a defective tape
drive would incorrectly report a tape mark on the read operation, followed
by an I/O error. This patch detects this error condition.
--------------------------------------------------------------------------------
Etrack Incident = ET257661
Description:
Improper initialization and handling of variables that manage open socket
connections was causing accidental closure of random sockets and file
descriptors. One of the symptoms, visible to end users, was an error
returned that indicated the user was not validated to access vmd when
attempting device configuration through the interface.
--------------------------------------------------------------------------------
Etrack Incident = ET361451 ET376485
Description:
SCSI protocol validation idiosyncrasies in the device driver for IBM's
PCI-X Dual Channel Ultra320 SCSI adapter prevented customers with tape
Libraries, with robotic control on SCSI LUN 1, from working correctly with
NetBackup. Attempts to obtain SCSI inquiry information from such devices
Returned a "SCSI adapter error, adapter status = 0x20" message.
--------------------------------------------------------------------------------
Etrack Incident = ET378009
Description:
The vmoprcmd -comment <drive_index> <comment_string> command would
overwrite the serial number specified if the comment_string size was 26.
--------------------------------------------------------------------------------
Etrack Incident = ET376808 ET308657
Description:
A vulnerability has been confirmed in the NetBackup Volume Manager
daemon (vmd). By sending a specially crafted packet to the Volume Manager,
a stack overflow occurs. This is caused by improper bounds checking.
Exploitation does not require authentication, thereby allowing a remote
attacker to take over the system or disrupt the backup capabilities.
Further testing and code inspection has revealed that all other
NetBackup 5.1 daemons are potentially affected in the same manner.
Therefore, any Master Servers, Media Servers, Clients and Console machines
at this version level are subject to this vulnerability. However,
NetBackup 5.1 database agents are not affected by this issue.
Refer to the Related Document section for more details.
--------------------------------------------------------------------------------
Etrack Incident = ET366838
Description:
The robotic inventory, vmupdate, may fail after encountering a media of
different density. The problem is that vmupdate was not taking the media
type of standalone media (if available) into consideration while
determining the media type of the media to be moved into the robot.
--------------------------------------------------------------------------------
Etrack Incident = ET396055
Description:
Directories were not being listed when browsing a backup on a policy that
allowed multiple data streams.
--------------------------------------------------------------------------------
Etrack Incident = ET296412 ET326949
Description:
On Linux platforms, NetBackup now skips the backup of the following new
file system types when ALL_LOCAL_DRIVES is specified: subfs, usbfs,
sysfs, rpc_pipefs, and binfmt_misc.
--------------------------------------------------------------------------------
Etrack Incident = ET339427
Description:
Fixed the following issues with bpcatarc and bpcatrm:
- The command was failing because of an incorrect recursion level for
image Query
- The backup job submitted by bpcatarc was failing when IDIRSTRUCT was set
to 1. Now it does not submit the job when IDIRSTRUCT is 1 and it provides
a message to the user.
- When the bpcatarc output was piped to bpcatrm, it would fail to remove
the .f files that where archived in that run.
- Images in linked folders where being archived.
--------------------------------------------------------------------------------
Etrack Incident = ET387412
Description:
This pack corrects a problem where corrupt TIR fragments can be written to
tape/disk during duplication when the reader/writer of the duplication are
on the same media server.
--------------------------------------------------------------------------------
Etrack Incident = ET355632
Description:
When bpbackup is run with the option to wait for the completion of the
backup job, a firewall between bpbackup and bprd can shut down the socket
while bpbackup waits for the completion status. This causes backint to
get stuck in an infinite loop during database backups.
--------------------------------------------------------------------------------
Etrack Incident = ET395007 ET401459 ET326517 ET326956 ET329213 ET378798 ET386527 ET387430 ET399373 ET209268
Description:
Added the following robots:
Falcon Stor
HP 1/8 autoloader
Sony CSM-20
NEC T16A
Updated robot entries for the following:
ADIC Scalar DLT 448
ATL P3000
Dell PV-132T
EXB-440
Spectralogic Python
Additional Notes:
Devices can be included in this file before they are officially supported.
--------------------------------------------------------------------------------
Etrack Incident = ET303673
Description:
Added support for IBM Ultrium 3 WORM drive/media.
--------------------------------------------------------------------------------
Etrack Incident = ET400302
Description:
Non-localized error messages are displayed when deleting Volumes.
--------------------------------------------------------------------------------
Etrack Incident = ET384760
Description:
A potential race condition exists if a Windows client backup receives a
134 status (resource busy).
The race condition was caused because the bpfis process had already started
taking a snapshot and bpfis has no knowledge of the resource failure. So,
the snap is taken, but when it is completed, there is no bpbrm and bpsched
(child) process running to accept the snap information.
When the job is later re-scheduled, this (unrequested) snap information was
passed back to the new bpbrm and bpsched. The unrequested snap information
confused bpsched (child).
--------------------------------------------------------------------------------
Etrack Incident = ET349768 ET405553
Description:
"bpdbm -consistency 2" rejects images that contain files belonging to
groups with very long names, which causes media to be reused before the
retention period expired.
Workaround:
When running bpdbm -consistency record the bad images, and then import
from the media these images were on.
--------------------------------------------------------------------------------
Etrack Incident = ET320044
Description:
bpdbm's host list was not consistent with bprd's host list. Bprd
was using FORCE_RESTORE_MEDIA_SERVER correctly, however, bpdbm was
not.
Workaround:
To avoid this issue, use the original media server name, or manually edit
the image database to refer to the alternate server.
--------------------------------------------------------------------------------
Etrack Incident = ET291762
Description:
Using the Windows NetBackup Backup, Archive, and Restore interface to
Restore files and folders with curly brace characters ('{' and '}') in
their names no longer fails with the error, "- no files matched in the
given date range."
Workaround:
The workaround is to use a wildcard specification that includes these files
with curly brace characters and then delete the files that you do not want
after the restore.
--------------------------------------------------------------------------------
Etrack Incident = ET325895
Description:
Error messages were being written to the bperror log and the bpdbm log
when TIR prunning was performed. Log messages says, "stat of <backup_id>
Failed".
Workaround:
Ignore the filesfilesize value in the image file.
--------------------------------------------------------------------------------
Etrack Incident = ET401486 ET403534
Description:
If the forked re-read storage unit process detected a transient error, it
would error, then exit and remove the scheduler message queues.
--------------------------------------------------------------------------------
Etrack Incident = ET292569
Description:
RMAN - alternate client restore times out after a successful partial
restore by media server.
Though this problem was encountered while doing an Oracle alternate client
restore, the nature of the problem is such that it can happen for any
XBSA-based partial restore.
When the client is done restoring - partial restore - but there is still
data to be read, the client sockets are closed but bptm/bpdm wait for the
data to be read causing it to eventually time out with an error.
--------------------------------------------------------------------------------
Etrack Incident = ET292569
Description:
RMAN - alternate client restore times out after a successful partial
restore by media server.
Though this problem was encountered while doing an Oracle alternate client
restore, the nature of the problem is such that it can happen for any
XBSA-based partial restore.
When the client is done restoring - partial restore - but there is still
data to be read, the client sockets are closed but bptm/bpdm wait for the
data to be read causing it to eventually time out with an error.
--------------------------------------------------------------------------------
Etrack Incident = ET404676
Description:
Bprd was dumping core, and the stack trace looked similar to the following:
Core was generated by `bprd -dontfork -mpxmain'.
Program terminated with a signal 11, “Segmentation fault”.
#0 0x00034ccc in write_monitor_status ()
(gdb) where
#0 0x00034ccc in write_monitor_status ()
#1 0x0003f204 in write_clnt_monitor_status ()
#2 0x0004f7e8 in terminate_restore ()
#3 0xfdd9efc8 in __tbl_10_small_start () from /usr/lib/libc.so.1
#4 0xfdd53990 in posixgetdst () from /usr/lib/libc.so.1
#5 0x0005f9ac in run_restores ()
#6 0x00063bd4 in run_mpx_main_bprd ()
#7 0x00021ad8 in main ()
--------------------------------------------------------------------------------
Etrack Incident = ET410095
Description:
The snap driver needs to be compiled on Solaris 10 to run on Solaris 10.
--------------------------------------------------------------------------------
Etrack Incident = ET410095
Description:
The snap driver needs to be compiled on Solaris 10 to run on Solaris 10.
--------------------------------------------------------------------------------
Etrack Incident = ET351797
Description:
Specifying ALL_LOCAL_DRIVES and Shadow Copy Components caused a
bpbkar32.exe crash.
Workaround:
Do not specify any shadow copy component entries in a file list with
ALL_LOCAL_DRIVES
--------------------------------------------------------------------------------
Etrack Incident = ET410877
Description:
** If using the command line bpubsora and bpubsdb2 would core dump
immediately upon execution.
If using the jbpSA interface, and the Oracle instance was clicked on, the
application would immediately exit with the message, "bin sh:6491 abort
status 134" and it would dump a core file.
--------------------------------------------------------------------------------
Etrack Incident = ET410877
Description:
** If using the command line bpubsora and bpubsdb2 would core dump
immediately upon execution.
If using the jbpSA interface, and the Oracle instance was clicked on, the
application would immediately exit with the message, "bin sh:6491 abort
status 134" and it would dump a core file.
--------------------------------------------------------------------------------
Etrack Incident = ET315094
Description:
FBU sfr restores would not start, and bprd was slow sorting the file
history.
Workaround:
To avoid these issues, do full a restore of the FBU sfr images to an
alternate location.
--------------------------------------------------------------------------------
Etrack Incident = ET371925
Description:
Because of security relates issues, the NetBackup 5.1 MP4 patch will be
released with the 3.23.59 version of MYSQL server binary.
--------------------------------------------------------------------------------
Etrack Incident = ET409436
Description:
When backing up a control file or archived logs using an Oracle
NAS_Snapshot policy to a disk STU, the backup would fail with a
status 20 - invalid command parameter. The following is an excerpt from
the dbclient log:
10:17:31.987 [15713] <4> check_comm_file_for_error: read comm file:<10:17:20
INF - Server status = 20>
10:17:31.987 [15713] <16> check_comm_file_for_error: ERR - server exited
with status 20: invalid command parameter
Workaround:
To avoid this issue, use a tape STU.
--------------------------------------------------------------------------------
Etrack Incident = ET405301
Description:
Eliminated a two-minute delay because of changes in the Red Hat 4.0 tape
driver.
--------------------------------------------------------------------------------
Etrack Incident = ET410720 ET410648
Description:
The NetBackup Volume Manager service, or Media Manager volume daemon, logs
messages as fatal when the message is actually informational and non-fatal.
In addition, some of the messages that were logged did not contain any
information about the problem, for example:
"terminating - unknown error -- -1 (-1)"
--------------------------------------------------------------------------------
Etrack Incident = ET403523
Description:
Changes were made to extended_copy.c broke compiles for media manager
binaries.
--------------------------------------------------------------------------------
Etrack Incident = ET369929
Description:
Bpbkar debug logs are currently limited to no more than 2 Gigabytes in
size. This change enables those logs to grow beyond 2 Gigabytes.
--------------------------------------------------------------------------------
Etrack Incident = ET413509
Description:
For a policy that uses wildcards (such as, *) in the file list and enables
multiple data streams, NetBackup will not include any dot files (such as,
.rhosts) at that directory level in the backup.
--------------------------------------------------------------------------------
Etrack Incident = ET336349
Description:
The NetBackup Client for NetWare will abend on a newly installed
NetWare OES server, if it has been configured to use the Novell Boot
Loader.
Workaround:
Update SYS:SYSTEM\BPSTART.NCF to include the -B switch, which manually
specifies the server load path. For example:
-bC:\NWSERVER
--------------------------------------------------------------------------------
Etrack Incident = ET411638
Description:
NDMP DAR restore fails when a tape has been imported
or ‘FORCE_RESTORE_MEDIA_SERVER’ is set to a different NetBackup platform
(UNIX to Windows or Windows to UNIX). The restore fails with the
following error:
''NDMP_LOG_NORMAL 0 RESTORE: FILE_NAME: File does not exist in backup.''
Workaround:
To avoid this issue, use the same platform (UNIX or Windows) where the
backup was performed.
Additional Notes:
The actual problem exists on the backup site. To do DAR restore using a
different NetBackup platform (UNIX to Windows or Windows to UNIX), you
have to perform a backup using this patch.
--------------------------------------------------------------------------------
Etrack Incident = ET375812
Description:
Large volume backups and restores for FlashBackup on Windows were not
functioning correctly because of memory allocation issues and 32-bit
integer overflow conditions.
--------------------------------------------------------------------------------
Etrack Incident = ET417525 ET415832
Description:
Security permissions of some temp files were 666.
--------------------------------------------------------------------------------
Etrack Incident = ET415713
Description:
Embedded wildcard characters (when a part of a multi-byte character in some
character sets) are processed when they should not be. A specific example
of this was sjis.
--------------------------------------------------------------------------------
Etrack Incident = ET416977
Description:
Added a definition for MNTTYPE_LO so lofs is properly handled.
--------------------------------------------------------------------------------
Etrack Incident = ET418808
Description:
If a customer created an image list (bpimage -create_image_list -client
<client) after an incremental TIR backup and then ran a full or cumulative
differential backup, the TIR_INFO field in the image was not updated to 10.
As a result, TIR information was not pruned from the backup image.
Workaround:
To resolve this issue, remove image list files from the client directory.
--------------------------------------------------------------------------------
Etrack Incident = ET419597
Description:
NetBackup can get an error 130 when backing up VxFS file systems on AIX
platforms which are mounted under non-VxFS file systems that have ACLs on
them.
--------------------------------------------------------------------------------
Etrack Incident = ET419976
Description:
The disks on Hitachi arrays at higher ports cannot be backed up when using
Shadowimage, because the CSH and Port number on the disks are more than two
digits long.
The fix enables ShadowImage to backup Tagma array disks that reside on
upper ports.
--------------------------------------------------------------------------------
Etrack Incident = ET418365
Description:
Activity Monitor failure for Red Hat 4.0 (no processes listed):
For Linux, the process examination code heretofore
assumed that all files /proc/*/stat were flat files that contained process
statistics. A read() on the stat file yielded salient process information.
However, read() will fail if stat is a >directory<, whereas open() will
not fail. On Red Hat 4.0, there is at least one case where stat
is a directory and not a file (/proc/net/stat). This fix skips the
read() of stat if it is not a regular file.
Additional Notes:
This issue appears to be limited to Red Hat 4.0.
--------------------------------------------------------------------------------
Etrack Incident = ET420858 ET422804
Description:
The first line of the header of the bpclimagelist command was occasionally
not displayed. This potentially could happen for any server type, but was
seen when the master server was Linux. This would affect BMR processing.
--------------------------------------------------------------------------------
Etrack Incident = ET422477
Description:
Staging from a disk staging storage unit that contained a space would
result in images that reside on the staging storage unit to not be
duplicated properly. However, the staging job would appear to succeed in
the Activity Monitor.
Workaround:
To avoid this issue, duplicate images to a new DSSU that do not contain a
space, and change any policies that write to original DSSU to use the new
DSSU.
--------------------------------------------------------------------------------
Etrack Incident = ET324507
Description:
backuptrace would ignore bpbrm lines when the backup is multiplexed.
--------------------------------------------------------------------------------
Etrack Incident = ET423255
Description:
A change has been made to avert a potential vulnerability in a Java
authentication service that runs on VERITAS NetBackup servers and clients.
This change prohibits remote attackers from executing arbitrary code on a
targeted system. In addition, Symantec recommends that users block the
affected ports from external network access.
For more information regarding this issue, refer to the following TechNote
on the Support Web site: http://support.veritas.com/docs/279085.
--------------------------------------------------------------------------------
Etrack Incident = ET427129
Description:
The new nbdbd 3.23.59 version, was causing VISD to go down on HP-UX
platforms.
--------------------------------------------------------------------------------
Etrack Incident = ET428603
Description:
In rare cases, SCSI inquiry in page code 83 (bytes 32 and 33) returns two
digits instead of four digits of numbers.
This fix filled in the digits and enabled Shadowimage to work.
--------------------------------------------------------------------------------
Etrack Incident = ET428603
Description:
In rare cases, SCSI inquiry in page code 83 (bytes 32 and 33) returns two
digits instead of four digits of numbers.
This fix filled in the digits and enabled Shadowimage to work.
--------------------------------------------------------------------------------
Etrack Incident = ET426052
Description:
Added client support for Windows 64-bit EM64T/AMD Opteron platform.
--------------------------------------------------------------------------------
Etrack Incident = ET497068
Description:
Added Windows 64-bit client type directories so that these new types appear
as client options.
--------------------------------------------------------------------------------
Etrack Incident = ET498525
Description:
In a rare condition, one or more backups could be corrupt when using Inline
Tape Copy (ITC) with multiplexed backups. This would potentially happen
under the following conditions:
- ITC was enabled with two or more active copies.
- Multiplexing was enabled with two or more active backups.
- The schedule configuration had "continue if the copy fails" flag
set for the copy that failed in the next step.
- A copy other than the first active copy encountered "end of media", so a
new media was needed for this copy. Then, there was a problem getting
or setting up media (for example, writing the media header) such that
this copy failed before continuing the backup.
- The last buffer written to the first active copy prior to the "end of
media" on the other copy was not the first active backup.
Workaround:
To avoid this problem, do one of the following:
- Edit the schedule configuration to disable the "continue if copy fails"
flag.
- Disable multiplexing.
- Disable ITC.
--------------------------------------------------------------------------------
Etrack Incident = ET429969
Description:
Binding to a socket that was already in use caused the following error to
appear in the bpsd log file:
10061 - WSAECONNREFUSED
Workaround:
To avoid this issue, use vnetd.
--------------------------------------------------------------------------------
Etrack Incident = ET500030
Description:
bpfilter did not take into consideration, the tar header (of type LF_BAD).
Because of this, multiplex restores would go past the end of the file and
end up encrypting everything that followed the file. Ideally, LF_END and
LF_BAD would be treated alike.
--------------------------------------------------------------------------------
Etrack Incident = ET500424
Description:
VxMS version 138 did not accept certain VSS snapshot names and caused
FlashBackups to fail. Version 140 of VxMS fixed this issue.
================================================================================
=========
NB_51_3_M
=========
Etrack Incident = ET262145
Description:
Some Linux file system types (such as, tmpfs, usbdevfs, devpts, and smbfs)
were being backed up unnecessarily.
Workaround:
To avoid this issue, manually exclude the file systems that are not to be
backed up.
(NetBackup Clients: Linux2.4 Linux64 )
--------------------------------------------------------------------------------
Etrack Incident = ET295113
Description:
The scheduler was modified to correctly process a status 134 error - race
condition in a multiplex backup.
This condition occurred when a multiplex child bpsched started a backup,
but the storage unit was over committed. A status 134 error was sent back
to the multiplex child bpsched, but the child schedule was already adding
another backup to the multiplex group.
The child schedule attempted to send the new jobs information to bpbrm, but
bpbrm had exited and the connection was defunct. The scheduler was
incorrectly logging the defunct connection in the job progress log. With
this change, the scheduler will not log the broken socket, but will
correctly read and process the status 134 from the in-bound connection.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET293094
Description:
When the SUI bit is set for some admin commands on an AIX 5.2 platform,
the system OS-level command will leave a directory in /tmp. This does not
occur for the root user.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET283809
Description:
Some levels of the Linux kernel do not support netstat correctly and the
message, "Device not found," is produced.
(All NetBackup Servers: Linux)
--------------------------------------------------------------------------------
Etrack Incident = ET270622 ET270620
Description:
There are cases when a NetBackup administrator can administrate most things
on a NetBackup master server (such as policy configuration), but is not
allowed to access Host Properties on clients.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276469
Description:
When attempting to quit by using the <ESC> key in bpadm, the key is
ineffective and it fails with the error, "Attribute 254 cannot be changed".
The <ESC> key works in AIX 5.1, so this issue seems to be exclusive to the
AIX 5.2 platform.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET270167
Description:
The SCSI TEST UNIT READY timeout value has been increased from 2 seconds to
60 seconds. RS6000 media servers running many concurrent jobs to multiple
drives simultaneously could result in SCSI TEST UNIT READY error messages
that are logged into the system log.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET294713
Description:
An problem occurred that caused bpretlevel loops to echo the word, "What?,"
when changing retention levels. No input seemed to be a satisfactory
response to exit the loop until a ctrl-C was entered.
Solution:
An earlier change had changed all occurrences of gets() to fgets(), but
failed to recognize that fgets() will return the '\n', whereas gets()
did not. The subsequent strcmp was not expecting the '\n', and
always failed. The solution was to strip the [new] trailing '\n' to
emulate the gets() behavior.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET265653 ET295655
Description:
Oracle Advanced Client FlashBackup backups did not work with the new
128-bit Encryption. The failures produced an error status 25 - cannot
connect on socket. The bpfis debug log included messages similar to the
following:
17:14:27.139 [13628] <16> bpfis main: FTL - snapshot creation failed,
status 25
17:15:19.842 [13628] <4> bpfis Exit: INF - EXIT STATUS 25: cannot connect
on socket
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET278777 ET272628
Description:
Database agent backups may fail if 128/256-bit Encryption is used and vnetd
"connect-backs" to the media server is used.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET278777 ET272628
Description:
Database agent backups may fail if 128/256-bit Encryption is used and vnetd
"connect-backs" to the media server is used.
All NetBackup Windows Clients (All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET276123 ET262341
Description:
Legacy Encryption may fail if the key file on the client is large. For
example, if 39 or more pass phrases are added to a key file on a Solaris
client, encrypted backups or restores are likely to fail. The error
message "bpbrm - read() or KEYREADFD failed" will typically be reported.
Workaround:
To avoid this issue, delete some pass phrases from the key file if they
are no longer needed.
((All NetBackup UNIX Clients))
--------------------------------------------------------------------------------
Etrack Incident = ET287818
Description:
The bpimmedia -spangroups output was not displaying a span group if the
span occurred in a copy other than copy 1, and one or more preceding copies
were expired.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET287614
Description:
A Synthetic Backup Image could be expired while it was still being written.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET210298
Description:
Backups of NDMP clients were not displaying intermediate directories when
browsing the backup contents. For example, if directory A/B/C was backed
up, /A and /A/B were not appearing when browsing the backup.
Although the NetBackup catalog contained the directory information, it did
not return it when the directory was not explicitly backed up. This fix
will enable browsing of these NDMP backups even if they were created prior
to the fix.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET297729
Description:
Oracle BLI backups failed on Solaris 10 when trying to backup multiple VxFS
partitions. The following would be logged:
For second or later file system, checkpoint mount may fail, bpbkar log:
11:58:55.199 [8353]<4> bpbkar set_clone_context: /usr/sbin/mount -F vxfs -o
ckpt=NetBackup_full_orac101_1106243854
/dev/dsk/c1t0d0s4:NetBackup_full_orac101_1106243854
/tmp/NetBackup_checkpoints/NetBackup_full_orac101_1106243854 >/dev/null
2>/dev/null
11:58:55.464 [8353] <16> bpbkar set_clone_context: ERR - Cannot mount
checkpoint NetBackup_full_orac101_1106243854 on
/tmp/NetBackup_checkpoints/NetBackup_full_orac101_1106243854.
(NetBackup Clients: HP HPIA64 RS6000_433 Solaris Solaris9 )
--------------------------------------------------------------------------------
Etrack Incident = ET296425
Description:
This bpbrm binary fixes an intermittent socket problem that occurred
during database backups. The backup would fail with a status of 54.
An examination of the bpbrm debug log showed the following type of message:
07:10:52.662 [3832.2768] <16> bpbrm listen_for_client: listen for client
timeout during accept from name listen socket after 60 seconds
An examination of the dbclient debug log showed the following type of
message:
02:24:07.956 [528.660] <8> connectSock: WRN - connect() to server failed,
No connection could be made because the target machine actively refused it.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET299437
Description:
An extended value (90 days, for example) for the "Keep Logs for x Days"
global configuration setting appears not to work. The debug logs are
still deleted after about a month on NetBackup media servers that
are included as part of the catalog backup configuration.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET272516
Description:
NetBackup sometimes incorrectly handles an exclude list with an asterisk
character "*" as an entry and an include list.
Workaround:
To avoid this issue, use either "/" or "/*" in the exclude list rather
than "*".
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET294004
Description:
Backing up files which do not have a group name can fail with a Status 13.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET291947
Description:
A SAP online backup is started and a Tablespace (PSAPES46CI) in the Oracle
database is placed in BACKUP MODE by the agent. However the backup receives
a STATUS 134 error in BPTM & BPBRM. The bpbkar, bptm, and bpbrm processes
exit and clean up after themselves. The SAP agent process remains active,
awaiting the backup to be queued and run again, leaving the Tablespace
(PSAPES46CI) still in BACKUP MODE.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET294365
Description:
vmd would hang after 21 days under VxSS/NBAC.
(All NetBackup Servers: Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET277025
Description:
When a local file system was auto-mounted to itself, NetBackup would
incorrectly skip the backup of the local file system on SGI systems. This
occurred when the hostname returned from the uname command and the
hostname stored in /etc/mtab were not both fully-qualified or both short
(for example, uname returns "machine" and /etc/mtab contains
"machine.domain.com").
Workaround:
Make sure the hostname returned by the uname command and the hostname
specified in /etc/mtab both use either the fully-qualified hostname or the
short name.
(NetBackup Clients: SGI65 )
--------------------------------------------------------------------------------
Etrack Incident = ET300673
Description:
If a /dev/rdsk device was used as the extended copy command target, the
code to convert the disk path to the sg path for scsi-passthrough commands
fails if the disk is managed by the ssd driver. This caused an error to be
logged in the bptm log this was similar to the following:
07:00:44.807 [13213] <2> setup_mover_tpc: open of pass-through path
/dev/rdsk/c1t50060E80034FED11d0s2 failed, unable to open
/devices/pci@1f,0/pci@5/SUNW,qlc@1/fp@0,0/sgd@w50060e80034fed11,0:raw, No
such file or directory
Workaround:
To avoid this issue, configure the corresponding sg path directly in
mover.conf.
(All NetBackup Servers: Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET299670 ET300174 ET300175 ET300176
Description:
bpbrm would hang if a database agent backup failed in a certain way.
The job would appear to be running in the Activity Monitor until an
administrator terminated it.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET301501
Description:
A Scheduler race condition was occurring: Child sched was sending a
‘done’ to the main sched while the main sched was sending a ‘term one’ to the
child sched. The child sched erroneously treated the ‘term one’ as the
response to the ‘done’ and then exited prematurely.
The correct child response is to discard (ignore) the ‘term one’
request and wait for the main sched response to the 'done' command.
Also missing was code to discard (ignore) a ‘suspend’ request after having
sent a 'done' to the main sched.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET300757
Description:
Interval-mode backup windows open longer than 6-days (144-hrs) resulted in
continuous backups on the seventh day, till the window closes.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET277470
Description:
Creating a long name in the hosts file and using that in a policy causes
various parts of BEDS to crash.
Workaround:
Ensure all names used to reference the host are 15 or fewer characters in
length
All NetBackup Windows Clients
--------------------------------------------------------------------------------
Etrack Incident = ET299481
Description:
ltid on Linux, when starting up vmd, causes the following to be logged
in /var/log/messages:
kernel: application bug: vmd(16703) has SIGCHLD set to SIG_IGN but calls
wait().
kernel: (see the NOTES section of 'man 2 wait'). Workaround activated.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET300679
Description:
Restoring compressed files from a FlashBackup on a Windows image resulted
in tar32 aborting when the master server was not a Windows platform.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET299843
Description:
Bptm, bpdm, or bpdbm would occasionally crash during TIR incremental
backups. Bprd, bpbrm, or bpsched were possibly affected as well.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET302767
Description:
The NetBackup Windows Administration Console will crash when attempting to
view certain jobs details in the Activity Monitor.
(All NetBackup Servers: Intel-NT Intel64)
--------------------------------------------------------------------------------
Etrack Incident = ET296072
Description:
When installing NetBackup Server on SUSE 9 systems, the installation
appeared to hang.
Additional Notes:
When the NetBackup 5.1GA install hangs, the make_scsi_dev process must be
killed from another terminal session. The NetBackup install should be
allowed to run to completion after make_scsi_dev has been killed. The
NetBackup 5.1MP3 Maintenance Pack must then be applied, and the
make_scsi_dev command run. Once make_scsi_dev completes, devices may be
configured.
(All NetBackup Servers: Linux)
--------------------------------------------------------------------------------
Etrack Incident = ET299846
Description:
Multiplexed backup jobs would occasionally hang. This would happen if the
bptm or bpdm parent process crashed, or if bptm could not successfully add
the TIR fragment for a TIR backup. It may also happen because of a timing
problem in bpbrm when new jobs are started.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET292021
Description:
A new bperror -jobgroupid option is introduced to search the error database
based on the job group ID. This is helpful for an Inline Tape Copy job
because it may log to multiple job IDs.
The syntax of the option is:
-jobgroupid job_group_id
where job_group_id is the job group ID number for the job.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET303111 ET301828 ET275764 ET293905 ET279264 ET284804 ET285117
ET270133 ET295671 ET295695
Description:
Added support for the following libraries:
EXABYTE VXA-2 1x10 Autoloader
EXABYTE 1x7 Autoloader
Added support for the following drives:
EXABYTE VXA-2
HP VS160
STK VDRIVE
Updated the following libraries:
COMPAQ TL895
DEC TL810
DEC TL895
HP MSL6000
HP MSL5000
HP UHDL
IBM 3583
IBM 4560SLX
IBM 7337
SONY PETASITE CSM200/100/60
Additional Notes:
Devices can be included in this file before they officially supported.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET293146
Description:
NetBackup processes may core if an error occurs. The stack from the core
file showed the processes were in the ndmp_user_scsi() or ndmp_user_tape()
functions.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET286656
Description:
bpverify would fail with a status 191 on a single master/media server
because of a socket timeout.
Workaround:
To avoid this issue, do not run bpverify on very large backups with a slow
or stressed network.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET266342
Description:
bplist returned duplicate listings for a DB2 file; and the .f file showed
only one instance of the file. This problem did not occur for other similar
client backups. If IMAGE_LIST file was removed, the file would be listed
only once.
Workaround:
To avoid this issue, turn off image indexing.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET303564 ET302345
Description:
The main sched failed to recognize a media server that was initially
off line, and then brought on line after the main sched went active.
If the media server was brought back while the scheduler was not active or
the master server was bounced, then everything worked correctly.
This fix enables the scheduler to ‘pick up’ the newly introduced media
server by using the ‘/bp/bin/admincmd/bpschedreq -read_stunits’ command.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET302345 ET303564
Description:
Backups using in line tape copy (ITC) and storage unit groups, failed to
give preference and select the local storage unit.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET303449
Description:
Backup failed with disk full even though there was still room on the disk.
Workaround:
The problem is the result of the settings in /etc/security/limits. The
value for file size could be increased to eliminate this problem.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET273869
Description:
CLUSTER_NAME is behaving the same way as if REQUIRED_INTERFACE was set.
Additional Notes:
This fix only applies to UNIX platforms. For Windows, the behavior has
always been as described. To turn off this behavior, add the following entry
in bp.conf:
ANY_CLUSTER_INTERFACE = 1
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET299846
Description:
bpdbm crashed due to a problem in a CPR case.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET304308
Description:
Staging/Duplication from a disk staging unit was not working when more
than 32765 files were in the staging area.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET270494
Description:
AIX would report the swap size in bytes and not kilobytes.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET309223
Description:
Backups were taking too long between job complete and 'end writing'.
To fix this problem, a correction was made to a recursive call in
searching for a storage unit.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET289123
Description:
Staging of images which were no longer in the catalog caused a 190 error.
Workaround:
Manually delete images which are not in the catalog.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET287633
Description:
When simultaneous, differential, incremental backups are done for the same
client, a subsequent synthetic backup for that client may fail.
Workaround:
To avoid this issue, do not run simultaneous backups on the same client.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET309223
Description:
Backups were taking too long between job complete and 'end writing'.
This problem was fixed by correcting a recursive call in searching for a
storage unit. In addition, removed a media server version check in the
main ITC - find storage unit loop.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET309085
Description:
On Solaris 10 platforms, NetBackup now skips the backup of the new file
system types ctfs and objfs.
Workaround:
Add the mount points of any objfs and ctfs file systems into the exclude
list on that client. For example, /system/object and /system/contract.
(NetBackup Clients: Solaris9 )
--------------------------------------------------------------------------------
Etrack Incident = ET321994 ET266342
Description:
While testing the NetBackup 5.1 MP3 patch, the ability to browse ASCII
catalogs did not exist.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET323188
Description:
Disk-staging of a directory with no images to stage causes the program to
terminate without informing its parent program that it has finished.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET323222
Description:
Disk-staging only stages the one copy of an image in the directory.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET320557
Description:
When a resume backup exits, it would attempt to send a request queue
message to the GUI. But the GUI does not read these messages, so the
message would stay queued until the main sched exited.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET323183
Description:
If the stream in control snap failed on a WOFB, then the other streams were
allowed to continue a non-snap backup. This was the correct operation. The
problem was that the non-control streams performed a backup of all files in
the file list, but they should have only backed up their specific files.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET316635
Description:
User backups failed to check if an exiting backup was the last backup job
from a specific SPID before removing all user backups from the work list.
This oversight caused the removal of active jobs from the work list, and
resulted in a sched core.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET323827 ET324060
Description:
DSSU stage to tape fails with a status 134 if the media becomes full.
A NON_MPX_RESTORE can cause bptm to select the wrong media during a restore
operation that spans tape.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET303490 ET315602 ET324213 ET323143 ET323736 ET277364
ET315274
Description:
Added the following library:
CERTANCE LTO3 CLL 6400
Updated the following library entries:
ADIC 10k
ADIC Scalar 24
ATL M2500
ATL M1500
Exabyte 480
IBM 3583
Additional Notes:
Devices can be included in this file before they are officially supported.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET315056
Description:
WORM tape support uses sixteen byte SCSI-3 commands. Some HBA's have
problems with these commands, and can cause SCSI Bus Reset's. A new touch
file, volmgr/database/NO_SIXTEEN_BYTE_CDB, can be created to disable these
commands. This will effect your ability to use WORM tape with some drives.
Solaris platforms have been found to not handle sixteen byte commands on
Parallel SCSI HBA's, so devices on these HBA's are automatically prevented
from using sixteen byte SCSI commands.
Workaround:
To avoid this issue, use an HBA that supports sixteen byte commands.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET384887
** Description:
ATTENTION: A related TechAlert exists for this issue.
NetBackup 5.1 Maintenance Pack 3 contained a condition under which data
loss might have occurred. This data loss would occur if the following
conditions were true:
NOTE: This issue only affects duplicate images using NetBackup 5.1_MP3.
The original or "source" image remains valid.
1. The original backup image must be written to tape. Virtual Tape
Libraries (VTL) will be affected by this as well. Disk-based backups are
not affected by this issue.
2. The original backup image must span more than one fragment. This would
be caused when multiplexing is enabled, when a backup image spans more than
one tape, when the maximum fragment size is reached, or when a checkpoint
occurs for checkpoint-restart backups.
3. The image must be a duplicate image and meet one of the following
conditions:
- The source image was not multiplexed on tape.
- The duplication is performed with the "Preserve Multiplexing" option
disabled.
If all the criteria above apply, then the second and subsequent copies of
backup images created by the duplicate operation will be written incorrectly.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
=========
NB_51_2_M
=========
Etrack Incident = ET218330
Description:
The synthetic backup timing was incorrect because to the time it took to
read the large extents.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET257821 ET232038
Description:
A large number of down storage units would result in and apparent scheduler
hang. In reality, the scheduler was timing out to each media server in
series. This fix recognizes and skips duplicate attempts to a down media
server.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET220801
Description:
"Ghost" backup jobs were appearing in the activity monitor. For example, a
deleted backup job could (partially) re-appear in the activity monitor.
This would only happen to a deleted job that was one number lower than a
multi-copy backup.
And example of this scenario is:
1) Job n finishes.
2) Job n is deleted from the activity monitor.
3) Multi-copy job n+1, n+2, n+3 finishes.
4) Activity monitor incorrectly updates jobs n, n+1, n+2.
5) Deleted job n (partially) re-appears.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET212797
Description:
The interrupt code paths for the Scheduler were simplified to eliminate
scheduler core dumps because the interrupt handlers were too "noisy".
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET211907 ET210846
Description:
If an in-line copy backup is multiplexed, and one of the copies gets a
failure (but continues), then subsequent additions to this troubled
multiplexed group generates excessive failure messages.
This code change inhibits adding new backup jobs to multiplexed groups
that contain a failing copy.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET218162
Description:
The bpplschedrec -deldayomonth command deletes a range of days instead of
the specified day.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET230710
Description:
A warning message is produced in RedHat 9 that indicates NetBackup is
issuing a wait after telling the operating system (OS) to ignore the
relevant signal. The OS disregards the wait so there is no actual
problem. This only affects bpcd in standalone mode.
Workaround:
The current solution for this issue is to ignore the warning message.
All NetBackup Windows Clients (All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET236426 ET230514
Description:
If NetBackup is configured to duplicate an image and also configured to
change the retention level of the new copy to a retention level that is
configured as "expires immediately", then the original image also appears
to be expired at the end of the duplication. Any backup taken after that,
returns the original image.
Workaround:
Any subsequent backup would expunge expired copy from the image file, and
bring back the original image (to image list and for restore operation)
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET234522
Description:
SCSI-3 library commands have standardized on op code 0x37 for the command
INITIALIZE ELEMENT STATUS w/RANGE. Previous implementations by numerous
vendors had used vendor-unique code 0xE7 - which was implemented by
NetBackup. SCSI-3 libraries will return sense data 20/00 when sent an E7.
Workaround:
To avoid this issue, use only tapes with barcodes that are valid for your
library vendor.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET234298
Description:
NetBackup was not properly using the firmware revision to retrieve the
correct device mapping in the device mappings file. This has never been a
problem in the field because VERITAS has never had to use the firmware
revision in the device mappings file - this is being done for new device
support.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET235646
Description:
Previously, cluster_config -a would go into an infinite loop when
configuring add-on products for monitoring when anything but a lowercase
"y" was entered at the following prompt:
You have chosen to monitor <addon>. Is this list correct? [y,n] (Y)
This issue has been resolved and the defaults, "Y" and "y" are all valid
responses.
Workaround:
To avoid this issue, you would have entered a lowercase "y" instead of an
uppercase "Y", or you could have taken the default.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET223542
Description:
The scheduler does not automatically start "bpexpdate -deassignempty" to
deassign media that do not have associated images after "bprecover -l" has
been run.
Workaround:
If a bprecover -l has been executed, remove the following file:
UNIX:
/usr/openv/netbackup/bin/deassign_lock
Windows:
VERITAS\NetBackup\Bin\deassign_lock
Do not remove this file if a catalog recover (bprecover -r) has been done.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET260510 ET259248
Description:
Only fails on a UNIX master. If a user archive fails to remove some files
(and generates a status 4), then the .f file is removed, and the files, up
to the failed file, are removed from the client.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET235410
Description:
It was possible that the bpjobd process may hang under certain conditions.
It has been demonstrated that this can happen if bpduplicate and bptm
processes are killed, but it may occur in other situations. The bpjobd
process does not die, it just loops consuming CPU time, and nothing is
written to the debug log. This prevented the job DB from being updated
with the status of existing jobs, or new jobs. bpjobd must be killed and
restarted by the system administrator to resolve this problem.
This problem is indicated by the following data in the bpjobd debug log,
referring to non-existent job ID 3:
18:45:00.734 [17433] <2> process_active_job: Frame type is 3 (jobid 3)
Sock (93)
18:45:00.734 [17433] <2> process_active_job: Adding provider(1) for job(3)
socket (93)
18:45:00.734 [17433] <2> process_active_job: Begin to write (JOBTRYFILE) at
offset 68 job (3)
Workaround:
To avoid this issue, kill and restart the hung bpjobd.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET223801
Description:
Under some circumstances, an image file might be expired on the Master
server and remain on the Media server. When media full is detected for a
DiskStaging storage unit, the media server attempts to remove the two
oldest images. If these images were no longer in the catalog on the Master
server, they could not be removed and the backup would fail.
The code was changed to remove the image files if the Master server no
longer had them in the catalog.
Workaround:
To avoid this issue, manually delete the image files that are causing the
problems along with their .ds file.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET221591 ET185355
Description:
The predict algorithm was not sorting correctly. It failed to position
cumulative incremental backups before the differential incremental backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET232060
Description:
To assist image database cleanup in 24x7 environments, image database
cleanup will always run in background. bpsched will not wait until the
cleanup is complete before scheduling new jobs. In addition, bpsched will
initiate an image database cleanup at periodic intervals, 12 hours by
default. The time interval may be changed by using the
bpconfig -cleanup_int <hours> command. For example, to change the interval
to 24 hours, execute the following command:
bpconfig -cleanup_int 24
Use a value of 0 to disable bpsched from starting the periodic image
database cleanup. The maximum value is 744 hours. Bpsched will still
perform an image database cleanup at the end of a session, however, it will
be run in background. On UNIX systems, bpadm may also be used to change
the interval.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET231652 ET224755
Description:
NetBackup was not properly examining the SCSI library mail slot status
bits; therefore, during ejects of more than a MAP-full of media, the eject
operation may abort after the first MAP-full is removed.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET257819
Description:
ltid will now terminate gracefully on UNIX platforms if it detects a
(invalid) device configuration where non-shared robots (robots with types
other than TLD, TL8, TLH and LMF) are configured on media servers with no
drives configured.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET232896
Description:
A problem exited causing a snapmirror-to-tape backup to fail with an
error 99.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET233382
Description:
In one instance, some catalog files were corrupted when the system ran out
of disk space. This caused bpdbm to crash whenever it tried to load the
corrupted catalog files. The command "bpdbm -consistency 2" also crashed
every time these corrupted files were checked.
The cause of this problem was the buffer overrun in libcatalog when loading
corrupted files: the pointers and string lengths are wrong after mapping
buffer overrun to fix the crash of bpdbm. In addition, code was added in
image.c to move corrupted catalog files out of /db/images (into /db.corrupt
directory) when running "bpdbm -consistency 2".
This new code resulted in the following new feature:
"bpdbm -consistency" will move corrupted catalog files out of the catalog
repository "/netbackup/db/images" and into "/netbackup/db.corrupt"
directory.
After running the above command, users will need to do the following:
1. Record all file paths in /db.corrupt.
2. Re-import the files in step 1.
3. Delete files in /db.corrupt.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET225198
Description:
The Phase II import of (non-binary) catalog files can fail because the
code tries to re-name an imported file before closing the file.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET221883
Description:
The file list for stream 1 is listed twice in the job details when stream 1
is force-started to take a snapshot. This error is only on multi-stream
snapshot backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET223787
Description:
The scheduler debug log was filling with the following message:
"storage unit was not assigned for clientjob <host name> <policy name>
<stunit name> <copy>"
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET260902 ET259243
Description:
When a user imports and/or restores a NetBackup 3.4 backup image in
NetBackup 5.0/5.1, files greater than 1GB in size fail to restore and
produce, "disk full at byte xxxxx", errors in the restore log. Image verify
also fails with a gigabytes size mismatch error for those files. In
addition, the Backup, Archive, and Restore GUI shows a discrepancy in
reporting the size of those files.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET258421 ET258293 ET261720 ET221167
Description:
When a database agent gets queued again, multiple times in a very short
period of time, there were cases where the server would time out waiting for
the client to connect. This fix prevents the client from getting behind the
server in these instances.
(NetBackup Clients: ALPHA HP Intel-NT Intel64 Linux2.4 RS6000_433
RS6000_51 Solaris )
--------------------------------------------------------------------------------
Etrack Incident = ET223261
Description:
Bpbrm can core dump if one of the signal handlers interrupts a malloc
or free when the code is updating the pointers that are used to track
the free or allocated space.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET264088
Description:
A multiplexed TIR duplication would fail if the prior backup duplication
had a TIR size greater than 1 Gigabyte, and would lead to a bptm crash.
Workaround:
Retry the duplication using non-multiplexed duplication. It is also
possible that the duplication will succeed by trying multiplexed
duplication.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET262983 ET262880
Description:
The backup job start time in the Activity Monitor was the same value for
each attempt. The start time value should be different on each attempt.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET192195
Description:
The ltid daemon will report operational issues at startup time using
localized error messages. In some cases, the boot-time console device can
not properly display non-ASCII characters, and can cause the boot-time
error conditions to be easily understood.
All of the error messages that can be presented at startup time are now
printed in ASCII only. No localization of the error messages will be
attempted by ltid.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET258458 ET167719
Description:
Added support for the following drive and media types for ACS:
Drives:
HP LTO-3
Media:
LTO-400G
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET263758
Description:
The scheduler would send a message on the wrong queue when it detected an
error on a client snapshot. This may result in a scheduler crash, if the
error happened frequently, or if the scheduler stayed running for a
prolonged period of time.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET265188
Description:
If an SSO drive is DOWN on all servers, and its scan host unregisters
(services are shutdown), vmd/DA will not transfer this drive to another
scan host. If the drive is brought UP on a server(s), it will have no
scan host, and will stay in the NO-SCAN state.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET265663
Description:
The image resume time limit was not always being updated on Checkpoint
backup jobs. Multiple attempts to resume a job might result in a backup
that is restarted instead of resumed.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET229709
Description:
The following bptm logging statement was confusing and has been changed.
09:37:46.035 [16536] <4> ndmp_internal_open_and_auth: NDMP server DEBUG
suggests protocol version 3
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET262921
Description:
The error "license use has been exceeded", status code 159, may have been
erroneously generated when using a SAN media server license for a media
server that is part of a cluster.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET212904
Description:
A failure occurred when attempting to restore from an image that was
backed up with the following setting:
"SET FILES = ( Multiple Files ) "
A status 5 would be generated.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET266252
Description:
There are two issues that may have caused backups to fail:
1. bpsched was not updating the RESUME_EXPIRATION time on failed backups.
2. bpdbm did not update the REQUEST_PID on subsequent Q_IMAGE_NEW requests.
Workaround:
To avoid this issue, make sure any job that has been resumed is finished
before the cleaning is done.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET233105
Description:
vmphyinv is unable to read the tape header in NDMP drives. This affects
NDMP drives and the NetWare media server (that are handled using NDMP
protocol).
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET233201
Description:
Some FlashBackup backups took longer on NetBackup 5.0 than they did for
NetBackup 3.4 because of the overhead of the binary catalog.
Workaround:
To avoid this issue, raise the maximum files-per-add value using the
/usr/openv/netbackup/MAX_FILES_PER_ADD to 9500. This will raise the limit
for all backups.
Additional Notes:
The number of files added per call to the catalog has been raised to 9500
from 500 for FlashBackup. This limit can be modified by putting the desired
value in /usr/openv/netbackup/FBU_MAX_FILES_PER_ADD, which only affects
FlashBackup backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET267530
Description:
Scheduler code was too inflexible and did not allow for easy addition of
new Windows Open File Backup (WOFB) snapshot clients. This change corrected
the scheduler and eased the addition of future Microsoft clients.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET263530
Description:
The method of configuring inetd has changed in Solaris10. Changed scripts
that deal with the modification of inetd.conf to use the new method if on a
Solaris 10 machine (Sparc or X86).
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET263530
Description:
The method of configuring inetd has changed in Solaris 10. Changed scripts
that deal with the modification of inetd.conf to use the new method if on a
Solaris 10 machine (Sparc or X86).
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET233198
Description:
bpSALinfo needs to be linked with a more recent set of SAL API libraries to
be able to interface to ccstor 4.0, and SPC 3.4. Minor changes are needed
to account for differences in return values for ccstor 4.0.
(All NetBackup Servers: HP RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET268568
Description:
The scheduler would take a long time scheduling multiple data stream jobs
if there were many policies that were due to be auto discovered at once.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET232056
Description:
When selecting a media ID for backup, if the allocation date in the media
database does not match the assigned date in the media manager database,
FREEZE the media because there could be database inconsistencies.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET268528
Description:
The cluster monitor script would not restart the stopped daemons on some
platforms. This happened because the .NetBackup_STATE file did not have a
NEWLINE at the end of the file, so SED did not see it as a complete line
and did not return a needed result.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET266303
Description:
For NDMP restores on NetApp filers, more than 1024 files can be restored
using DAR.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET269452
Description:
This version of the SG driver adds support for Solaris 10.
(All NetBackup Servers: Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET263450
Description:
Errors occurred when the Sony TSL-11000 tape library would put the
cleaning tape away to its appropriate slot.
Workaround:
The cleaning actual works, but the cleaning statistics do not get updated.
Running a tpclean -M <drive name> after the failed clean will correct this.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET269768
Description:
Add support for RedHat 3.0 on the Zseries platform. This required a new
client type which is called IBMzSeriesLinux2.4.21. The number is based off
the operating system (OS) kernel.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET269768
Description:
Add support for RedHat 3.0 on the Zseries platform. This required a new
client type which is called IBMzSeriesLinux2.4.21. The number is based off
the operating system (OS) kernel.
(NetBackup Clients: new )
--------------------------------------------------------------------------------
Etrack Incident = ET233695
Description:
During multiplexed restores, when multiple tar instances are restoring
files onto the same directory tree, some of the files may not get restored
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET264130 ET266195 ET265296 ET262877 ET258096 ET223049
ET258202 ET258252 ET257553 ET230175 ET223889 ET223643
Description:
Added support for the following libraries:
- ATL M2500 (fixed inventory issue)
- Certance CLL3200 (fixed TLD(AVR) issue)
- HP DAT72X6
- IBM 3623 2SX
- IBM DDS4
- NEC T40A
- Sony Lib-162
- Sony Petasite CSM200/100/60 (serialization change with 4.01)
Added support for the following drives:
- Benchmark VS640
- Sony SDX-900V
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET269914
Description:
BLI failed to mount a clustered storage checkpoint. The bpbkar log
displayed the following:
14:56:41.914 [13792] <4> bpbkar set_clone_context: /usr/sbin/mount -F
vxfs -ockpt=NetBackup_full_test_1096401397 /dev/vx/dsk/cfs_dg/rb_test
NetBackup_full_test_1096401397 /rb_test_s/NetBackup_checkpoints
NetBackup_full_test_1096401397 >/dev/null 2>/dev/null
14:56:42.409 [13792] <16> bpbkar set_clone_context: ERR - Cannot mount
checkpoint NetBackup_full_test_1096401397
on /rb_test_s/NetBackup_checkpoints/NetBackup_full_test_1096401397.
(NetBackup Clients: HP RS6000_433 Solaris )
--------------------------------------------------------------------------------
Etrack Incident = ET269541
Description:
An extra invocation of a function was resetting the delay time on tape UN
mounts for user backups. Removal of the extra invocation should produce
more optimal UN mounting of tapes when performing user backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET269768
Description:
Added support for IBMzSeriesLinux2.4.21.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET235418
Description:
In certain configurations, calls to gethostbyname (3nsl) may fail. In such
cases, VERITAS did not force a DNS lookup for a fully qualified host name.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET270921
Description:
Because of a race condition, ACSLS drives may get downed during mount
processing.
Additional Notes:
The signature of this race condition is that the event viewer (or syslog)
will note errors in semaphore operations from ACSD, and then take the
drive down.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET274324
Description:
When multiple disk staging jobs are initiated and the DSSU is at or
near capacity, some jobs may fail with a status 84 error, "disk full,
attempting to remove the same staged images".
BPDM logs contain entries that indicate no matching entries found in the
catalog, (db_IMAGE: db_end_sts() failed: no entity was found).
Workaround:
If the DSSU is nearing capacity, manually remove the previously staged
images. These images are identified by an additional zero-length file of
the same name, with a ".ds" extension.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET274137
Description:
A bug in write_data_tir() had corrupted the TIR file when the disk was full
and when writing to a disk staging unit.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276040
Description:
Maintenance Pack 1 for the NetBackup 5.1 release introduced a performance
issue with NDMP DAR restores for catalogs with a large number of files.
This fix corrects the performance problem.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276089
Description:
Added 64-bit Oracle support for Linux on the AMD64 and EM64T platforms.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276089
Description:
Add 64-bit Oracle support for Linux on the AMD64 and EM64T platforms.
(NetBackup Clients: Linux2.4 )
--------------------------------------------------------------------------------
Etrack Incident = ET272137 ET274040
Description:
Corrected a case where incremental backups were backing up complete data
sets (full backup). This happened only for multi-stream WOFB or advanced
client where a previous non-control stream backup failed.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET273464
Description:
Corrected a problem where tapes would not unmount in a reasonable amount of
time.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET271482
Description:
Re-naming a schedule, and then attempting an immediate backup using the
newly named schedule caused a core dump in the scheduler.
Workaround:
To avoid this issue, do not add, remove, or modify policy or schedule
information and then attempt an immediate backup while scheduler is
running.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET275271
Description:
In a rare circumstance, the last block of a multiplexed, duplicated copy of
a TIR backup may be corrupt. A verification of the copy would not be
successful. The original backup is not affected. This does not affect
non-multiplexed duplications of a multiplexed TIR backup.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276969
Description:
Corrected a problem that allowed an incorrect block size to be put in tape
backup headers. This problem occurred when the file SIZE_DATA_BUFFERS
exists to change default tape block size, a backup is duplicated to tape
and the source/destination media is on the same machine. The incorrect
block size in the tape backup headers would not prevent restores from
working, however, it would prevent import phase 2 from working if that
copy (with the incorrect backup header) was ever imported.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276830
Description:
User archive or synthetic backups that have checkpoint enabled, and receive
a 134 (resource busy) will be rescheduled. But the Activity Monitor lists
the job as "done" status 134.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET272608
Description:
If a set of configuration options in bp.conf is very large, it can cause
bpgetconfig and the GUI to not display any of the items. This applies to
items in bp.conf with the same name, for example CONNECT_OPTIONS or SERVER.
Additionally, if that set of configuration options is edited in the GUI,
for example, to add a new server to the CONNECT_OPTIONS, all of the
previous data for that set of options will be lost and only the new data
will be in bp.conf.
Workaround:
To avoid the issue, reduce the number of configuration options with the
same name.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET273254
Description:
NDMP restores fail with Hitachi NAS when restoring the entire backup image.
The NDMP log shows the path being sent twice in the NDMP_DATA_START_RECOVER
message.
Workaround:
Restores will work as long as the file system path is not selected (the
path that was initially backed up). If the entire image is needed for
restore, select all the files and directories in the file system path
instead.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET272608
Description:
If a set of configuration options in bp.conf is very large, it can cause
bpgetconfig and the GUI to not display any of the items. This applies to
items in bp.conf with the same name, for example CONNECT_OPTIONS or SERVER.
Additionally, if that set of configuration options is edited in the GUI,
for example, to add a new server to the CONNECT_OPTIONS, all of the
previous data for that set of options will be lost and only the new data
will be in bp.conf.
Workaround:
To avoid this issue, reduce the number of configuration options with the
same name.
All NetBackup Windows Clients All NetBackup PC Clients
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET258096 ET270045 ET268600 ET260459 ET273679 ET220083 ET260459
ET274370
Description:
Added support for the following libraries:
- ARCHIVE Python 06241-XXX910B
- Certance CLL3200
- Diligent VTF Open
- IBM DDS4 Autoloader
- IBM 3623 2SX
- SEAGATE DAT 9SP40-400910B"
Added support for the following drives:
- HP LTO3
- IBM Ultirum-TD3, 3580-TD3
Additional Notes:
Devices can be included in this file before they officially supported.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET259704
Description:
When running Windows Open File Backup enabled backups using VSS on
Windows 2003 clients, performance of the backup is considerably slower
compared to when running a Windows Open File Backup enabled backup using
VSP or with Windows Open File Backup disabled. Only backup performance is
affected and no data loss occurs as a result of using VSS as the Windows
Open File Backup snapshot provider.
All NetBackup Windows Clients
--------------------------------------------------------------------------------
Etrack Incident = ET230111
Description:
When backing up encrypted data on Windows 2000 clients using Windows Open
File Backup to backup an open/active file, the encrypted data will be
inaccessible to the backup, and the backup will end partially successful.
Additional Notes:
In NetBackup 5.1 MP2, Windows Open File Backup snapshots will be disabled
when backing up encrypted data on Windows 2000 clients. This is not
applicable to Windows operating systems newer than Windows 2000 such as
Windows XP and Windows 2003. A logging message similar to the one below may
appear in the bpbkar32 debug logs indicating this on Windows 2000 clients:
4:29:09.328 PM: [2620.2648] < 4> file_access::V_OpenForRead: INF -
Disabling
snapshots for: D:\lock\ORW_SN.txt
By disabling snapshots for Windows encrypted data on Windows 2000 clients,
these files may not be accessible to the backup if they are busy or locked
by another process. If Windows Open File Backup protection is important,
Windows encryption should be disabled on the files being backed up.
All NetBackup Windows Clients
--------------------------------------------------------------------------------
Etrack Incident = ET271139
Description:
Image fragments are sometimes left behind on the disk storage unit after
the image expiration has occurred, and the accumulation of such "rogue"
fragments can eventually create backup failures unless the rogue fragments
are manually deleted by an administrator.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET279250
Description:
Backups of encrypted files completed with a "partially successful"
message, however, the encrypted files were not backed up.
(NetBackup Clients: Intel-NT )
--------------------------------------------------------------------------------
Etrack Incident = ET283490
Description:
Initiating multiple Oracle backups caused crashes in scheduler. Each of the
Oracle backups attempted to read/update/write a group ID file (used for
synching snap backups), but there was no lock mechanism on the file access,
so each Oracle backup had the potential to corrupt another’s file access.
When this problem was discovered, the environment was running about
20-Oracle backups (one policy per client).
Workaround:
To avoid the issue, stagger window openings for Oracle backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET279063
Description:
When performing duplication and direct shared memory was being used instead
of sockets, increase the timeout the writing bptm/bpdm waits for the
reader to setup the shared memory. Previously, if the system was
very busy, the timeout was not enough.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET277797
Description:
Robot inventory update was not taking into account the media type for
existing standalone media that have been moved into a robot. It was only
looking at the barcode rules. This was the basic cause of the error "media
type and volume group mismatch (101)". Robot inventory update now updates
the media residences to be robotic without changing the media type.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET284896
Description:
When running a Restore job, the information about the job does not appear
in the Job Details -> Job Overview tab in the Activity Monitor, however,
the job still completes successfully.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET287099
Description:
After running FlashBackup backups that were aborted, a snapcachelist
command showed the snapshot cache device still in use even though the
snaplist command did not return any snapshots. Eventually, the snapshots
would fail because of the 32 snapshots per cache device. A reboot is
required to recover the use of the cache device.
Additional Notes:
The snap driver has been modified to use a different flag bit to denote a
snapshot in creation versus a snapshot that is being deleted. A snapshot in
creation is allowed to be listed by snaplist, and can be deleted.
FlashBackup has been changed to remember the snapshot ID between the time
that the snapshot is initially created to the time that the snapshot device
is opened so that it can be deleted.
(NetBackup Clients: Solaris Solaris9 )
--------------------------------------------------------------------------------
Etrack Incident = ET287015
Description:
When running backups with Windows Open File Backup and running multi-
streamed and multiplexed backups, some backups may end with status
41, "network connection timed out". When these errors occur, there should
be multiple (at least 30) log statements such as the one below in the
bpbkar32 logs for the same bpbkar32 process:
12:04:07.788 AM: [4008.3816] <4> V_Snapshot::V_Snapshot_TimeoutCheck: INF -
Attempting to query volume snapshots: bpfis query -id 1103259068
12:04:08.788 AM: [4008.3816] <4> V_Snapshot::V_Snapshot_Destroy: INF -
Attempting to destroy volume snapshots: bpfis delete -id 1103259068
12:04:11.788 AM: [4008.3816] <4> V_Snapshot::V_Snapshot_Destroy: INF -
Snapshot Destroy EXIT STATUS 0: the requested operation was successfully
completed
This is not an issue for non multi-streamed Windows Open File Backup jobs
or backup jobs that have Windows Open File Backup disabled.
Workaround:
To avoid this issue, disable Windows Open File Backup for the client that
have backups end with status 41.
All NetBackup Windows Clients
--------------------------------------------------------------------------------
Etrack Incident = ET287831
Description:
The snap driver has a limit of 32 active snapshots per cache device, which
is not enough in some cases.
Workaround:
To avoid this issue, reduce the file list to use less than 33 file systems,
and create more policies that use different cache partitions.
Additional Notes:
The limit has been increased to 64.
(NetBackup Clients: Solaris Solaris9 )
--------------------------------------------------------------------------------
Etrack Incident = ET284489
Description:
Tar could get into an infinite loop while attempting to restore hard links
when the linked file does not exist or is not a part of the restore job.
Workaround:
To avoid this issue, restore the actual file as well as the hard link, or
make sure the file the hard link is linked to exists for the restore.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET286770 ET216480 ET286770 ET287990
Description:
A VERITAS Volume Manager (VxVM) backup with instant recovery fails when two
file systems are listed. You should not truncate the snapvol name in
pfi_create_snapshot_vol().
(NetBackup Clients: HP Intel-NT Solaris )
--------------------------------------------------------------------------------
Etrack Incident = ET287990 ET216480 ET286770 ET287990
Description:
The following list identifies a few issues with FlashSnap FIM if full-sized
instant snapshots (FMR3 style) are used.
A. You should not use vxsnap split before a disk group split; it deletes
DCO so that FMR will not be utilized.
B. Advanced client does the refresh of the first volume only before a disk
group split; it causes 'fsck' to fail on the alternate client.
C. The calculation of tmp_vol string len is incorrect, resulting in not
enough space being allocated which then causes a core dump.
(NetBackup Clients: HP Intel-NT Solaris )
--------------------------------------------------------------------------------
Etrack Incident = ET288026
Description:
The .f and non-.f files were not getting deleted on a NAS snapshot client.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET289149
Description:
Corrected possible skipped backups if the frequency is a multiple of
24-hours (interval mode) and the backup window is greater than 24-hours.
Only the first backup was run in the failure scenario.
Workaround:
To avoid this issue, make the backup windows no greater than 24-hours.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET289704
Description:
The XBSA Agent PFI backups resulted in status 156 errors. In this
failure, the scheduler is receiving a file list, ?start?(FI_FILE_START)
and file list ?end?(FI_FILE_END), however, it would not receive a file
list ?body?.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET292496 ET289149
Description:
Corrected possible skipped backups if the frequency is a multiple of
24 hours (interval mode) and the backup window is greater than 24 hours.
Only the first backup was run in the failure scenario.
Workaround:
To avoid this issue, make the backup windows no greater than 24-hours.
(All NetBackup Servers)
================================================================================
=========
NB_51_1_M
=========
Etrack Incident = ET185139
Description:
The drives used for disk staging backups are not counted correctly. This
results in under utilization of drives for other backup operations.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET207814
Description:
When scheduler verbose is set greater than 0, there is the potential for
jobs to exit with a status 50. This is more likely to occur on systems that
experience connection issues. The situation occurs when the scheduler
prints a message and loses the errno value. Armed with the incorrect errno
value, the scheduler exits (status 50) the current running job.
Workaround:
Set scheduler verbose to zero.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET206000
Description:
A correction to potential scheduler problems with memory leaks and
non-initialized malloc space has been made.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET205980
Description:
The restart of a frozen image backup job failed. The issue was found using
the following setup:
- VxVM snapshot method to backup a VxFS/VxVM file system.
- checkpoint every 5 minutes.
- scheduled frequency is 1 hour.
- backup attempts are set 2 tries per 1 hour.
The following was the attempted failure scenario:
- Launch the backup and wait for a checkpoint to be taken.
- killed bpbkar processes to put the job into an incomplete state.
- waited for the job is restarted (1 hour later).
- the attempt fails.
- bpfis is attempting to create a new frozen image.
- the previous frozen image still exists.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET207493
Description:
When using the bpexpdate -recalculate or the bpexpdate -backupid commands
to extend the expiration date of one or more backups that have multiple
copies, the media expiration may not be correct for copies on removable
media if the first copy is on disk.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET207498
Description:
You may encounter an occasional failure when duplicating True Image Restore
(TIR) backups. The failure may occur if there is a delay in mounting the
write media.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET167637
Description:
With STK's new version of SSI, certain environment variables are required
to be set when the vm.conf entry ACS_TCP_RPCSERVICE is set. The variables
are set to a default behavior unless they are set differently in the
vm.conf.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET185363
Description:
Corrected compile warnings in the bpadm storage unit free and print
schedule calendar invocations.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET185146
Description:
Bpduplicate, during disk staging duplication, prints the following message
even though no error has occurred.
11/22/2003 11:25:35 AM - Error bpduplicate(pid=1636) from host dcdell1,
COMPLETE_MOUNT QUANTUMDLT70000 0
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET209936
Description:
Resume of a single-stream, suspended job would backup the complete policy
file list instead of only the stream's portion of the file list.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET184689
Description:
In cases where a master was set to require NBAC and a client was
set to automatic without a credential HP-UX would not return the
same error code as other platforms. In such cases error codes 192
and 193 would be documented in the debug log, but would be exposed
back to the user as error code 23.
Workaround:
To avoid this issue, set the bp.conf entry USE_VXSS = PROHIBITED or
USE_VXSS = REQUIRED on client system.
(NetBackup Clients: HP )
--------------------------------------------------------------------------------
Etrack Incident = ET185135
Description:
bpexpdate would not notify the user that they were unable to use bpexpdate
because of access control limiting their access. Even though the binary
returned the appropriate error level (such as, 117, 116, or 118) a human-
readable message was not returned.
Workaround:
To avoid this issue, manually check the error level that bpexpdate returns
(for example, echo %ERRORLEVEL% or echo $?).
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET185137
Description:
Even though the NBU_Operators group did not have backup permission on
Policies, they were still able to perform manual backups. The NBU_Operator
group should be able to do manual backups, however, it should always
require the Backup permission on Policies.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET206796
Description:
Using the options that allow a UNIX file system to be restored without
crossing mount-points fails when restoring all files under the path
instead of just the single file system.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET210609
Description:
A backup with TIR and CPR both enabled in the policy attributes, and files,
instead of directories, specified in the policy file list, can fail with a
status of 130 (signal 11 in the bpbkar log), if the backup goes on for more
than the checkpoint interval.
Workaround:
To avoid this issue, do not enter files in the policy file list, with TIR
and CPR enabled in the policy attributes. TIR works on directories, and so
specifying files in the policy file list is not of much use.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET184717
Description:
Bptm or bpdm may crash after bpsched restarts following a bpsched crash
and subsequently resuming checkpointed backup jobs on media servers. This
change cancels existing duplicate jobs when this situation occurs, or
prevents the crash from occurring.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET210606
Description:
The TLH robotics does not wait for the library auto drive cleaning to finish
before an error occurs and passes off to bptm.
Additional Notes:
In addition, two new syslog entries have been added.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET184971 ET187776
Description:
The VOPIE Enhanced Authentication may not work when host names are greater
than 32 characters. You may see Authentication Failed (error code 160)
messages when connecting with systems with host names or client names
greater than 32 characters.
All NetBackup Windows Clients (All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET184975
Description:
When bpinst -ENCRYPTION pushes the bpkeyutil binary to a UNIX client, the
mode bits of the binary are 0500 on the client. The bits should be 0555.
This could cause a problem if the NetBackup administrator is not running
as "root".
Workaround:
To avoid this issue, login as root on the client and enter this command:
chmod 555 /usr/openv/netbackup/bin/bpkeyutil
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET211098
Description:
The tldcd and tl8cd parent did not check for unit attentions every
15 seconds and caused the AUTO_UPDATE_ROBOT feature to not working properly.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET211433
Description:
When "Browsing for restore", the files that ended with an asterisks '*'
or an "at sign" '@' would have these characters truncated. This prevented
those files from being restored.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET185147
Description:
The following compiler warnings indicate that part of the stack is being
overwritten on these calls.
warning C4133: 'function' : incompatible types -from 'unsigned int *'
to 'time_t *'
Workaround:
This is a rather broad problem that has not appeared in testing, rather as
warnings during a compile. Without these changes, the user should verify
the output of bpcatlist and also verify that bpcatarc is working by
checking the CATARC value after the .f file is archived (of any backup
image that is archived).
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET208248
Description:
Performing backups using an exclude or include list with non-fully-
qualified path names could cause significantly slower backup performance.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = RSVmn50653
Description:
Synthetic Backup fails if the "image copy 1" has been expired, even if a
secondary copy exists.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET213216
Description:
Systems with multiple tld or multiple tl8 robots may see strange behavior
across robot numbers on remote (not control hosts) media servers when the
daemon is bounced. In the worst case, this can include the drives going
into "AVR" control mode on the noncontrol host media servers. Generally,
this is started by a catastrophic failure of one of the robots.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET212553
Description:
Vault currently does not give the user the ability to specify which host
is selected for use as the "robotic control host" for an API robot.
Workaround:
Hand editing the Vault Configuration file
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET209681 ET207797
Description:
Using the command, bpimage -cleanout -client dill-bk, to reduce size of a
database can cause bpimage to core dump.
Workaround:
To avoid this issue, allow the nightly clean-up to clean the catalog.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = RSVmn49450
Description:
NBU_Operators did not have the backup permission on Policies, yet they were
able to perform manual backups. The NBU_Operator group should be able to do
manual backups, but it should always require the Backup permission on
Policies.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET184689 ET229056
Description:
In cases where a master was set to require NBAC and a client was set to
automatic without a credential, HP-UX would not return the same error code
as other platforms. In such cases, error codes 192 and 193 would be
documented in the debug log, but would be exposed back to the user as
error code 23.
Workaround:
To avoid this issue, set the bp.conf entry on the client system to:
USE_VXSS = PROHIBITED or USE_VXSS = REQUIRED
(NetBackup Clients: HP )
--------------------------------------------------------------------------------
Etrack Incident = RSVmn50953
Description:
XBP was inserting and deleting items from lists one item at a time making
the browse for restores very slow. This was done because of deficiencies in
some X-Servers' support of 1.1 Motif. This is not an efficient way to
handle lists with a large number of items.
With X-Servers that support Motif 1.2 this is not an issue. The code was
changed to delete all items at once and to insert multiple items at the
same time.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET164190 ET164186 ET208178 ET217630 ET218069 ET219861
ET220083 ET223049 ET217198
Description:
Fixed inventory issues for the following libraries:
DELL PV-132T, IBM 3582, Sepaton S100, All STK Libraries, HP A5617A
The added libraries are:
Certance CLL3200
EMC Disk Library
STK SL500
IBM 3590H11
IBM 36232LX
The added drives:
Sony SDX-400V
Added Sony SDX-500V
Certance Ultrium 3
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET213982
Description:
Corrected a restore problem from disk that resulted in a status 92:
- Image contained multiple disk fragments and first file of the
restore was not in the first fragment.
- The size the fragments prior to the one containing the first
file were not a multiple of 1K.
The above conditions can occur during a multiplexed multiple-copy
backup, one or more copies to disk, and one or more copies to tape.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET217243
Description:
Adjusted a loop used during the deletion of a multiplexed group. This
loop was not functioning correctly on the Windows 2000 platform and
resulted in the removal of additional multiplexed groups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET218075
Description:
Bprd hits a segmentation fault and core dumps. The issue appears to occur
when the log file path for the bprestore log exceeds 58 characters. This
only occurs when the media server involved is not the master server, and
when FQDN names are used.
This issue was resolved by increasing the command buffer size to
BUFFLEN (4096), and implemented bounds checking on all string buffers,
where practicable, via the use of V_strlcpy, V_strlcat, and snprintf().
Additional Notes:
While libVcom.so contains no changes, it is required to resolve new entry
points introduced in libsfr.so.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET217923
Description:
Added Solaris10 and Solaris_x86_10 to drop-down menu.
Additional Notes:
For "Solaris Solaris10" to show when configuring clients into a backup
policy, you must execute the following command:
/usr/openv/netbackup/bin/goodies/new_clients
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET206031 ET211885ET211834
Description:
Resolutions to the following issues have been added.
ET206031 Memory leak encountered in the XBSA library.
ET211885 DB2 BLI backup failure
ET211834 BLI backups with multiple streams fail
All NetBackup Windows Clients HP Linux2.4 Linux64 RS6000_433 RS6000_51 SGI
Solaris Solaris9
--------------------------------------------------------------------------------
Etrack Incident = ET211885
Description:
Changes have been added to support DB2 BLI backups.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET219207 ET229830
Description:
Added Solaris10 and Solaris10 x86 to the client drop-down list.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET221820
Description:
There is a potential for missed files if all the following conditions
exist:
1) Advanced client backup.
2) Incremental backup.
3) Multi-stream backup.
4) Backup stream is not stream 1.
Files modified after the snapshot is taken but before the (non-stream 1)
image is created may be missed.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET219245
Description:
After an NDMP import, the file count drops by a very large number.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET225034
Description:
An incremental multi-stream snap backup with forced stream 1 can miss files.
The problem exists when the following conditions are true:
- Snap backup (advanced client, WOFB).
- Incremental backup.
- Multi-stream backup.
- Stream 1 is not due, but is force-started to perform the snapshot
operation. Aggravated by ?frequency? mode (frequency is NOT a multiple of
24-hours).
The problem can be minimized/eliminated by configuring for an "all stream"
backup:
- Use calendar-based scheduling.
- Use ?interval? mode (frequency is multiple of 24-hours).
- Avoid multi-streaming snap backups.
This problem occurs because of an oversight in cloning a due stream into
a non-due stream1. Remember, stream1 is not due, but is required to
perform the snap shot operation. The due stream information is copied in
mass to create the stream1 information, and then adjusted to make it
stream1 unique. The error occurs because the backup times (last backup,
last full backup, last incremental backup) were not updated to the correct
stream1 values. As a result of this, stream1 uses the stream2 image
creation times. Image creation times are used to compute the delta time
which determines the targeted incremental backup files.
Unfortunately stream2 also suffers the same failure. The stream2
incremental backup should be based upon the stream1 snap time, but stream1
is using the incorrect time for the snap. So, all the due streams suffer
this failure.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET219966
Description:
Resume of a checkpoint backup could select a different media server if a
failure on initial backup attempt occurred.
The failure scenario is:
1) Backup reached a checkpoint.
2) Then failed (this case was a 13 "read failed").
3) Scheduler attempts to programmatically re-submit.
4) But the drives were re-set & downed.
5) So, the job exited with a 213 "no storage units available".
The original storage unit information was "lost" so the job could
potentially resume on a different media server. If resumed in this way,
the backed up data would reside on two different media servers and any
subsequent restore would fail.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET229834
Description:
A loss of multiplexed groups within bpsched.exe resulted in the failure to
cancel multiplexed backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET229949
Description:
Unable to restore using individual file restores through FlashBackup on a
Windows client with an AIX master server, if a zero length file exists in
the restore list.
Workaround:
Restore will work if no zero length file is included in the restore list.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET228739
Description:
Listing the master in the ClientList no longer causes update_clients to
fail. To correct this issue, a check was added in bpinst so that if the
client being shipped to is the master, and to not try and ship the java tar
ball, which is very large. It will also corrupt the file, and it is
already in the right location. In addition, a change was made to
update_clients in the case where the master is the client as nbj.conf has
moved to nbj.conf.bak.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET229844
Description:
An intermittent problem with installer no longer occurs when trying to
install to remote clients. The following two files were not being copied
to %TEMP%\X86, and caused the following error to appear, "The system
cannot find the file specified" on Install GUI.
VERITAS Maintenance Pack.msi
Data1.cab
Workaround:
To avoid this issue, do the following:
1) Start the installer. Select "Select from available computer on the
network". Click Next.
2) Copy VERITAS Maintenance Pack.msi Data1.cab to the %TEMP%\X86 directory.
(easiest way to find what %TEMP% is, is to open a cmd shell and run echo
%TEMP%, usually its something like C:\Documents and
Settings\<UserName>\Local Settings\Temp\X86).
3) Proceed with install.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET230865
Description:
An initialization failure against nbwin.exe is generated when a
non-administration account initiates a restore. A pop-up is generated
stating that nbwin.exe failed to initialize 0xc0000022. A change to the
permissions was made to msvcr71.dll so that the user "Users" has Read and
Execute permissions.
Workaround:
Manually change the security permissions on msvcr71.dll to allow Read and
Execute permissions for "Users" user.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET211802
Description:
The cancel_all_exceeded_max_wait() function does not scale well in large
environments. This change modifies the way incomplete jobs are processed
when they have exceeded the maximum incomplete time.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET224678
Description:
Improved VMD debug logging to better enable analysis for slow servers.
Allow control of VMD child timeouts, such that slow servers can be
interrupted, before VMD encounters connection timeout's on other
connections.
Add "VMD_CHILD_TIMEOUT = xx", to the vm.conf file. The value "xx" is the
number of seconds to wait for a child to release its lock on the volume
database. Allowable time is 60 - 360 seconds, where 360 is the default
(and value used prior to this patch).
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET218653
Description:
A race condition existed when two restore jobs needed the same piece of
media, from the same server, when this media was not already in the media
database. The condition could cause the media to be added twice and one
of the restores to fail.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET231394
Description:
Oracle RMAN backups taken using UNIX Media servers fail with a status 25.
(All UNIX NetBackup Servers)
===============================================================================
Requirement: NB_CLT_51_4_M
================================================================================
This Maintenance Pack provides fixes for VERITAS NetBackup (tm) UNIX servers.
NetBackup UNIX Add-on products, Database Agents and Java GUI have separate
Maintenance Packs.
Symantec recommends that the backing up of active file systems be avoided,
or the use of snapshot technologies be implemented. The directory structure
reported back from the file system (to NetBackup) may not contain all of the
files available during the time of backup. NetBackup will not report errors
in many cases where the file's existence is not known to NetBackup as reported
by the file system.
================================================================================
=================
PACK DEPENDENCIES
=================
-- NB_CLT_51_4_M_<6 digit number>.tar must be installed before this
Maintenance Pack is installed.
-- Installation of this Maintenance Pack requires version 1.19.4.23 of
the Vrts_pack.install script.
I. DOWNLOAD INSTRUCTIONS
II. KNOWN ISSUES
III. INSTALLATION INSTRUCTIONS
IV. UNINSTALL INSTRUCTIONS
V. DESCRIPTION OF PROBLEMS FIXED
Current Pack
NB_51_3_M
NB_51_2_M
NB_51_1_M
=========================
I. DOWNLOAD INSTRUCTIONS
=========================
1) Download the NB_CLT_51_4_M_<6 digit number>.tar and
NB_51_4_M_<6 digit number>.<server>.tar files into the
/tmp directory,
where <6 digit number> is an internal tracking identifier
where <server> is alpha_5, hp_ux, linux, rs6000, sgi, solaris
NOTE: NB_CLT_51_4_M_<6 digit number>.tar has the client binaries and
NB_51_4_M_<6 digit number>.<server>.tar has the server binaries and BOTH
must be installed.
2) Extract the NB_CLT_51_4_M_<6 digit number>.tar and the
NB_51_4_M_<6 digit number>.<server>.tar files.
tar xvf NB_CLT_51_4_M_<6 digit number>.tar
tar xvf NB_51_4_M_<6 digit number>.<server>.tar
NB_51_4_M will create the files:
VrtsNB_51_4_M.README
VrtsNB_51_4_M.<server>.tar.Z
VrtsNB_51_4_M.postuninstall
VrtsNB_51_4_M.postinstall
VrtsNB_51_4_M.preinstall
Vrts_pack.install
NB_CLT_51_4_M will create the files:
VrtsNB_CLT_51_4_M.README
VrtsNB_CLT_51_4_M.tar.Z
VrtsNB_CLT_51_4_M.postuninstall
VrtsNB_CLT_51_4_M.postinstall
VrtsNB_CLT_51_4_M.preinstall
==================
II. KNOWN ISSUES
==================
Description:
Solaris 10 is now supported as follows:
Hardware Support
---------------------------------------------
SPARC NetBackup Server and Client
Intel x86 NetBackup Client
The above support is for base OS support only. With NetBackup 5.1 MP2,
the "Zones" or "Containers" feature introduced in Solaris 10 is not
supported. NetBackup is supported on a system pruning global zone
only, any additional local zones or containers are not supported at
this time.
The support of NetBackup Client for Solaris 10 x86 on AMD Opteron
servers is not supported with NetBackup 5.1 MP2, but the support may
be announced at a later time based on successful test results.
Use of NetBackup Advanced Client methods are supported on Solaris 10
beginning with VERITAS Storage Foundation Suite version 4.1 releasing
in 2005. Check the Sun/VERITAS web-site for availability.
Solaris 10 is not supported on the base CD-ROM version of NetBackup 5.1.
There are known connection and Java GUI issues that will be encountered
if you attempt to run the NetBackup 5.1 GA version on Solaris 10. This
is due to a new inetd design method introduced in Solaris 10.
Therefore, this NetBackup 5.1 MP2 patch update must be applied.
The following script must be run after this patch is installed in order
for "Solaris Solaris 10" or "Solaris Solaris_x86_10" to show as a client
selection in the drop down list for backup policies:
/usr/openv/netbackup/bin/goodies/new_clients
Upon running this script, Solaris 10 choices will be available in the
drop-down menu for Solaris 10 clients.
===============================
III. INSTALLATION INSTRUCTIONS
===============================
For Maintenance Pack installation on a UNIX Cluster Environment:
NOTE: Click on the "Download Now" link, near the bottom of this document
prior to running the following installation procedure for this pack.
1) Ensure that prior to installing the maintenance pack, NetBackup is at
release level 5.1 and configured to run in a cluster.
2) Freeze the NetBackup group (This will avoid a 'failover' during a patch
installation).
3) Install this Maintenance Pack on the inactive node(s) of the cluster
(follow steps 1-3 below).
4) Install this Maintenance Pack on the active node of the cluster (follow
steps 1-3 below).
5) Unfreeze the NetBackup group.
--------------------------------------------------------------------------------
There are two ways to install the client maintenance pack software.
1. Remote Installation: Loads the software on a master server with the intent
of pushing client software out to affected clients.
2. Local Installation: Loads and installs the software only to this local
machine.
Remote client install:
As root on the NetBackup Master Server:
1) This pack contains a full release of the IBMzSeriesLinux 2.4.21 client. To
install, you must first create the following directory:
mkdir /usr/openv/netbackup/client/Linux/IBMzSeriesLinux2.4.21
If you intend to install the Encryption libraries for this client type,
you must create the following directory:
mkdir /usr/openv/netbackup/crypt/Linux/IBMzSeriesLinux2.4.21
2) Install NB_51_4_M and NB_CLT_51_4_M Maintenance Pack binaries.
cd /tmp
/bin/sh Vrts_pack.install
3) Restart daemons.
/usr/openv/netbackup/bin/initbprd
/usr/openv/volmgr/bin/ltid -v
4) Update the NetBackup clients with the update_clients script.
/usr/openv/netbackup/bin/update_clients <hardware> <os>
where <hardware> <os> is one of the following:
ALPHA OSF1_V5
HP9000-700 HP-UX11.00
HP9000-800 HP-UX11.00
HP9000-700 HP-UX11.11
HP9000-800 HP-UX11.11
Linux IBMzSeriesLinux2.4
Linux IBMzSeriesLinux2.4.21
INTEL FreeBSD4.5
Linux RedHat2.4
MACINTOSH MacOSXS10.2
RS6000 AIX4.3.3
RS6000 AIX5
SCO UnixWare7.1
SGI IRIX65
Solaris Solaris7
Solaris Solaris8
Solaris Solaris9
Solaris Solaris10 *
Solaris Solaris_x86_7
Solaris Solaris_x86_8
Solaris Solaris_x86_9
Solaris Solaris_x86_10 *
* - For "Solaris Solaris10" or "Solaris Solaris_x86_10" to show when
configuring clients into a backup policy, the following must be
executed on the server:
/usr/openv/netbackup/bin/goodies/new_clients
Note: The /usr/openv/netbackup/bin/update_clients command without
any parameters will update all the UNIX clients.
Note: When updating an RS6000 client, there may be circumstances
where update_clients will fail with an error similar to
this:
Couldn't open /usr/openv/lib/libVmangle.so on client
Client open errno = 26
If this happens, execute /usr/sbin/slibclean on the client
to be updated and re-run update_clients.
If the client (CLT) .Z file and README exist in the installation
directory during the installation of the server maintenance pack,
the Vrts_pack.install script will install the client maintenance
pack automatically. The client maintenance pack will NOT be
installed automatically during a reinstall of the server
maintenance pack.
Additional Notes:
If non-root administrators use the GUI only, the nonroot_admin
script no longer needs to be run. If the non-root administrators
use the command line or bpadm, the group and file permissions
will have to be changed manually on the NetBackup binaries.
Users can write their own script. The script is being phased
out because there is a slight security risk that non-root users
may be able to execute NetBackup commands only because those users
are part of a group that is allowed to execute NetBackup commands.
For "Solaris Solaris10" to show when configuring clients into a backup
policy, the following command must be executed:
/usr/openv/netbackup/bin/goodies/new_clients
Local client install:
The install script will determine if a local client install is appropriate
and choose the appropriate client type to install.
As root on the NetBackup client:
1) Install NB_CLT_51_4_M Maintenance Pack binaries.
cd /tmp
/bin/sh Vrts_pack.install
NOTE: It is not possible to install the new Linux client
(LinuxIBMzSeriesLinux 2.4.21) locally. It must be installed on the
Master server and pushed to the client.
===========================
IV. UNINSTALL INSTRUCTIONS
===========================
Note: This will ONLY uninstall the Maintenance Pack from your local machine.
1) Close the NetBackup user interfaces.
Make sure the NetBackup server has no active jobs running (for
example, backups, restores, or duplications).
If a database agent is being used, such as Oracle,
ensure that the database services are stopped.
2) Change directory to the patch save directory.
Substitute the pack name for ${PACK} in the following command:
cd /usr/openv/pack/${PACK}/save
3) Run the un-install script:
./Vrts_pack.uninstall
4) Verify that the pack uninstalled successfully by checking
/usr/openv/pack/pack.history.
5) If update_clients was run after the pack was originally INSTALLED,
run it again after that pack is successfully UNINSTALLED.
6) If necessary, restart the NetBackup and Media Manager daemons:
/usr/openv/netbackup/bin/goodies/netbackup start
=================================
V. DESCRIPTION OF PROBLEMS FIXED
=================================
The following are descriptions of the problems fixed.
Please read the entire document before installing.
README Conventions:
Description
Describes a particular problem contained in this Maintenance Pack.
** Description **
Describes a problem that can lead to potential data loss. Please
read these problem descriptions carefully.
Workaround
Any available workarounds to a problem are also listed. Workarounds
can be used INSTEAD of applying the patch, however, Symantec strongly
recommends the "best practice" of being at the latest patch level.
Additional Notes
Any additional information regarding this problem or feature is included.
=============
Current pack
=============
================================================================================
Etrack Incident = ET342058
Description:
If a machine has NetBackup and a Maintenance Pack (MP) installed, and an
uninstall is performed of the MP that requires a reboot, and NetBackup is
uninstalled before the reboot, the DLL named msvcp70.dll is removed. The
installer has been modified to ignore MFC DLL in the Windows system32
folder.
Workaround:
To avoid this issue, perform a reboot when you are requested after first
uninstalling the MP and before you uninstall NetBackup.
--------------------------------------------------------------------------------
Etrack Incident = ET263530
Description:
For Solaris10, a change needed to be made to make sure that /etc/services
was updated before inetd.conf. This was only an issue on UNIX servers and
only if YP was not configured.
--------------------------------------------------------------------------------
Etrack Incident = ET312135
Description:
NetBackup 5.1 clusters need to be aware of NetBackup Operations Manager.
Without this patch NetBackup Operations Manager will not work in a
NetBackup 5.1 cluster.
Workaround:
The user could add the following lines to
/usr/openv/netbackup/cluster<cluster_platform>/<platform>_NBU_RSP.
DIR=var mkdir
DIR=var/global mv
LINK=var/global
--------------------------------------------------------------------------------
Etrack Incident = ET374817
Description:
A user currently cannot enter more than 99 dates into a calendar
schedule. To resolve this issue, increase the size of some of the arrays.
--------------------------------------------------------------------------------
Etrack Incident = ET354621
Description:
DSSU relocation would fail after upgrading to NetBackup 50_MP5 or 51_MP3.
--------------------------------------------------------------------------------
Etrack Incident = ET341834
Description:
Disk Staging cleanup was not occurring, nor was it notifying the user
quickly enough if the Disk Staging Unit was a NAS mounted file system.
--------------------------------------------------------------------------------
Etrack Incident = ET346373 ET366137 ET322959
Description:
Corrected the display of new ACS vendor media types T10000T1, T10000TS,
and T10000CT. In addition, corrected the allowable types that ACS vendor
media type LTO_400G.
--------------------------------------------------------------------------------
Etrack Incident = ET361923
Description:
Some character sets do not support upper case even though the underlying
language does. This caused some wildcard constructs to fail because the
upper and lower cases were not distinguishable.
Workaround:
To resolve this issue, the wildcard constructs were reworked so [] and '-'
could not be used with upper and lower case.
--------------------------------------------------------------------------------
Etrack Incident = ET325076
Description:
Backups of NetLink 2.x and higher files failed.
--------------------------------------------------------------------------------
Etrack Incident = ET311173 ET223766
Description:
If the add_slave_on_clients command is unable to add the media server
entries to one client, it may incorrectly update the media servers on
subsequent clients. Typically, only one media server will be added to
the remaining clients.
Workaround:
To avoid this issue, perform one of the following two work-arounds:
- Use the Host Properties interface to update the server list for
the clients.
- Use the bpsetconfig admincmd to update the server list for the clients.
--------------------------------------------------------------------------------
Etrack Incident = ET364429
Description:
** A stack trace showed that bpbrm was in the exit() system call, causing
bpbrm to dump core. This does not affect backups or restores.
--------------------------------------------------------------------------------
Etrack Incident = ET336219
Description:
The removal of a copy of a backup image was not occurring when the copy
expired. This occurred when the operating system was Linux and the file
system (where the NetBackup database resided) was VxFS.
--------------------------------------------------------------------------------
Etrack Incident = ET355614
Description:
If a backup is being run on an NFS file system and the NFS server is down,
the backup would hang.
--------------------------------------------------------------------------------
Etrack Incident = ET342043 ET207782
Description:
Restoring a large number of Windows files to UNIX would lead to being
overwhelmed with .SeCuRiTy files. (These files are from the Windows
security data that is invalid for UNIX.)
In addition, a Windows to UNIX restore did not indicate success because
of a failure to restore the Windows security data to UNIX.
Workaround:
To avoid these issues, do not restore Windows files to UNIX, and if you do
attempt to do this, make sure you clean up the .SeCuRiTy* files.
--------------------------------------------------------------------------------
Etrack Incident = ET352808 ET343561
Description:
Reverse host name lookup is required for outgoing NetBackup connections.
With this patch, outgoing connections from NetBackup servers that are not
configured for NBAC will not require reserve host name lookup.
--------------------------------------------------------------------------------
Etrack Incident = ET366163 ET355610
Description:
In certain firewall configurations, multistream backups from a Linux
client may fail with a status 25 if the vnetd connect-back is used.
--------------------------------------------------------------------------------
Etrack Incident = ET326978
Description:
After getting a, "client hostname could not be found" error, communication
between bpbrm and bpsched would fail. The bpbrm log contained the following
messages:
18:08:51.752 [24593] <2> put_strlen_str: cannot write data to network: Bad
file number (9)
18:08:51.752 [24593] <16> bpbrm send_parent_msg: could not write
COMPLETE_MOUNT V02107 0 message to stderr
--------------------------------------------------------------------------------
Etrack Incident = ET316059
Description:
Resumed restore from a number of images that involved a number of tapes
fails with the following error messages in the bprd log:
01:27:54.837 [557110] <2> put_length_bytes_nonblocking: cannot write data
to network: There is no process to read data written to a pipe. (32)
01:27:54.837 [557110] <2> write_files_to_active_sockets: Could not write
to file list socket 9 - connection dropped or not connected. Errno = 32:
There is no process to read data written to a pipe.
--------------------------------------------------------------------------------
Etrack Incident = ET341332
Description:
If using Multiple Data Streams, there would possibly be a long delay in
scheduling jobs. The bpsched log would show many hosts_equal calls that
were contributing to the delay.
--------------------------------------------------------------------------------
Etrack Incident = ET330685
Description:
Canceling a queued Vault job caused all backups to cancel.
--------------------------------------------------------------------------------
Etrack Incident = ET321736
Description:
The use of a closed socket error message was filling the scheduler log.
--------------------------------------------------------------------------------
Etrack Incident = ET322851 ET322817
Description:
When a server includes multiple network interfaces and is configured for
NBAC, other NetBackup machines may reject connections from the server
with authorization failures. This patch insures that the server presents
the correct credential to the other NetBackup machines.
--------------------------------------------------------------------------------
Etrack Incident = ET336943
Description:
A user backup could not be cancelled using the Activity Monitor.
--------------------------------------------------------------------------------
Etrack Incident = ET347554
Description:
On Windows platforms only: The scheduler would enter an infinite loop after
A read message queue failure occurred.
--------------------------------------------------------------------------------
Etrack Incident = ET349783
Description:
The scheduler would enter an infinite loop if all of the following
occurred:
- The backup is checkpoint enabled.
- The backup is twinning.
- One of the twined copies failed.
- The non-failed copy exits with a failure status.
--------------------------------------------------------------------------------
Etrack Incident = ET310386
Description:
Previously, there was a logic error that prevented customers from using
cluster_config -r to add robots for monitoring. This patch fixes the error.
Workaround:
The avoid this issue, add the name of the robotic daemon directly to the
PROBE_PROCS line in the <platform> _RSP file.
--------------------------------------------------------------------------------
Etrack Incident = ET363305
Description:
Adding a job to a multiplex group failed to check if a HOLD_134 exists on
the multiplex storage unit. This resulted in excessive 134 status backups.
--------------------------------------------------------------------------------
Etrack Incident = ET366178
Description:
Field 11, of the 'bpdbjobs -all' output, did not reflect the correct end
time of the last try.
--------------------------------------------------------------------------------
Etrack Incident = ET314632
Description:
A change has been made so the scheduler will no longer dump core.
--------------------------------------------------------------------------------
Etrack Incident = ET325883
Description:
The Activity Monitor would continuously display 'positioning' for
multi-fragment backups.
--------------------------------------------------------------------------------
Etrack Incident = ET316646
Description:
The scheduler performance suffered if the media server connection is
intermittent. The problem is aggravated when schedule continuously polls
the media server for NetBackup version information.
--------------------------------------------------------------------------------
Etrack Incident = ET352026
Description:
If an attempt to create a binary .f file is done with file entries without
user or group data, the attempt would fail but leave a zero-length .f
file. This caused a subsequent attempt to create the .f file in ASCII
format.
--------------------------------------------------------------------------------
Etrack Incident = ET315664
Description:
If you attempt to perform a cat_convert on an image where the policy name
contains a period (or full-stop) and the letter [f] follows immediately
after the period (or full stop), then the cat_convert says the file does
not exist.
Workaround:
To avoid this issue, do not create policies with names that contain ".f".
--------------------------------------------------------------------------------
Etrack Incident = ET311559 ET316451
Description:
Backup jobs with multiple streams where the individual stream selections
start with the same path can produce image files which, when browsed, do no
list all of the files in the backup.
Also, .f files created in NetBackup 4.5 versions for non-standard,
non-Windows clients may not show all files in the backup when browsed
under NetBackup 5.1.
--------------------------------------------------------------------------------
Etrack Incident = ET315663
Description:
Policy names containing ".f" cause premature EOF on Full TIR Backups on
the first try of jobid and when following an incremental.
Workaround:
To avoid this issue, do not use policy names that contain ".f".
--------------------------------------------------------------------------------
Etrack Incident = ET375771
Description:
If many Flashbackup images are created using an ASCII image format,
cleanup fails with the message, "too many open files".
Workaround:
To avoid this issue, use the binary .f file format to create the
Flashbackup images.
--------------------------------------------------------------------------------
Etrack Incident = ET371664
Description:
The synthetic backup process would begin reading the wrong block number
from tape during the synthetic backup image creation, causing the
synthetic backup to fail. This would happen when there was more than
one tape drive available to read the component images.
The following sample log message is a symptom of this failure:
bptm 12090 first XFERBLOCK indicated block number of 478341120 but initial
positioning was for block 193352448
This message indicates an, “unexpected block number read,” will be seen in
both the job details for the synthetic backup job, and in the bptm log.
Workaround:
To avoid this issue, down drives so that only two drives are available for
creating the synthetic backup - one drive for reading and one drive for
writing.
After the synthetic backup job mounts the first volume to be read and
mounts the volume to be written, you can safely increase the number
of “up” drives.
--------------------------------------------------------------------------------
Etrack Incident = ET380235
Description:
If the NOexpire touch file was set, then the NetBackup database image
cleanup was disabled. However, media would still expire. The media no
longer expires if the NOexpire touch file is set.
Bpexpdate -deassignempty, which deassigns media if there are no
corresponding images in the image database, is disabled.
The NOexpire touch file is set in the following locations:
For UNIX:
/usr/openv/netbackup/bin/NOexpire
For Windows:
VERITAS\NetBackup\Bin\NOexpire
--------------------------------------------------------------------------------
Etrack Incident = ET355414 ET499608
Description:
Synthetic Backup would produce an image that contains only the last
Incremental if a previous full synthetic backup (from incremental backups)
was manually expired.
--------------------------------------------------------------------------------
Etrack Incident = ET314637 ET405575
Description:
"bpdbm -consistency" failed to move out corrupted image files in
"catstore". It also failed to move out the zero byte image files
(which should have been handled by cleanup but it also failed).
Additional Notes:
On UNIX platforms, “bpdbm consistency” failed to move out the corrupted
image files if /usr/openv/netbackup/db and
/usr/openv/netbackup/db/images/<client> were not on the same volume. In
such case, the corrupted image files will be left to their original
locations.
--------------------------------------------------------------------------------
Etrack Incident = ET379404
Description:
The C_GET_HOSTINFO bprd request (from bpgetconfig -g for example) would
fail if there was not a cipher file on the client/host requested.
To avoid this issue, make sure a non-empty string is passed between the
dbmisc APIs, bprd, and bpgetconfig for the cipher list even if encryption
is not configured on the client.
--------------------------------------------------------------------------------
Etrack Incident = ET365011
Description:
When a C_GET_HOSTINFO request is made from a pre-NetBackup 6.0
installation, the [expected] request type and client name are appended to
the MAGIC string. For NetBackup 6.0 and later, these two fields are sent
separately. Infer the level of the requestor by counting the MAGIC string
arguments, and retrieve the type and client name accordingly. Once
inferred, ensure that the 'acks' that are sent to the requestor are in
synch with the requestor’s expectations.
--------------------------------------------------------------------------------
Etrack Incident = ET378815
Description:
The scheduler predict was not functioning correctly. The input date was
decremented so that it never exceeded the current date.
--------------------------------------------------------------------------------
Etrack Incident = ET371647
Description:
An XBSA application would hang during a shared memory backup (such as, the
client and the media server are the same machine) if a problem occurred
that caused the bptm to exit. The BSASendData() call did not return.
--------------------------------------------------------------------------------
Etrack Incident = ET340730
Description:
If the bpgetmedia options -h and -M were both specified, the -h argument
would always inherit the -M argument, and the original -h value would be
lost.
--------------------------------------------------------------------------------
Etrack Incident = ET323264
Description:
bprd could not clean a vault session directory that contains files larger
than 2 Gigabytes.
Additional Notes:
To resolve this issue, use lstat64() for LARGEFILES platforms in the
recursive_dir_delete() function. Used primarily by bprd for periodic
deletion of log files (and vault session files)... and by bpdbm for
removing expired back-ups.
--------------------------------------------------------------------------------
Etrack Incident = ET386569
Description:
Schedules that specify "instant recovery" should not have backups that
multi-stream. It is possible that a policy may try to multi-stream in
the following circumstances:
- Frozen image enabled.
- PFI enabled.
- Multi-streaming enabled.
- Policy contains schedules with instant recovery enabled.
--------------------------------------------------------------------------------
Etrack Incident = ET313352
Description:
Disk staging only submits one duplication job at a time so it was unable
to make use of idle tape drives.
Additional Notes:
Up to four bpduplicate jobs are submitted at a time. The duplication jobs
contain images grouped into jobs of 25 Gigabytes. By creating files named
MAX_STAGING_JOBS and STAGING_JOB_KB_LIMIT in the NetBackup directory on the
master server, these defaults can be adjusted as needed. For the
STAGING_JOB_KB_LIMIT, the value is in kilobytes and the default is 26214400
(25 gigabytes). A smaller value causes more jobs to be submitted, but
there will only be MAX_STAGING_JOBS running at a time.
--------------------------------------------------------------------------------
Etrack Incident = ET374768
Description:
ACS robot inventory may timeout when ACSLS server is overloaded.
--------------------------------------------------------------------------------
Etrack Incident = ET392051
Description:
Checkpoint-enabled user backups were performing an automatic retry
(limited by the failure history). But user backups should not auto retry.
--------------------------------------------------------------------------------
Etrack Incident = ET373913
Description:
The TPC selection process does not allow for selecting parallel mover
devices when available to improve performance and when multiple backups
are active.
Additional Notes:
This path adds a "GROUP" keyword in mover.conf. The GROUP keyword has a
parameter that is the name of a file in /usr/openv/volmgr/database that has
a list of mover devices that have the same capability. Each bptm that
executes the GROUP keyword will select the next device in the file in a
round robin fashion. A separate file is maintained to keep track of the
next device index.
--------------------------------------------------------------------------------
Etrack Incident = ET374770 ET388367
Description:
Tlhcd may dump core when running dump_tables from ltid -tables command.
Tlhcd needs to check for a media ID with all spaces for that ID to be the
last media in the list that is returned from the robot controller.
--------------------------------------------------------------------------------
Etrack Incident = ET394946
Description:
A restore from a FlashBackup image hung on a Linux server with an
indication of a double free of a memory allocation.
--------------------------------------------------------------------------------
Etrack Incident = ET394942
Description:
User backups or archives that do not find an active policy/schedule to use
were marked as resumable, but without the policy/schedule information, the
job could not be resumed.
--------------------------------------------------------------------------------
Etrack Incident = ET394948
Description:
A duplication may have been reported as successful if a defective tape
drive would incorrectly report a tape mark on the read operation, followed
by an I/O error. This patch detects this error condition.
--------------------------------------------------------------------------------
Etrack Incident = ET257661
Description:
Improper initialization and handling of variables that manage open socket
connections was causing accidental closure of random sockets and file
descriptors. One of the symptoms, visible to end users, was an error
returned that indicated the user was not validated to access vmd when
attempting device configuration through the interface.
--------------------------------------------------------------------------------
Etrack Incident = ET361451 ET376485
Description:
SCSI protocol validation idiosyncrasies in the device driver for IBM's
PCI-X Dual Channel Ultra320 SCSI adapter prevented customers with tape
Libraries, with robotic control on SCSI LUN 1, from working correctly with
NetBackup. Attempts to obtain SCSI inquiry information from such devices
Returned a "SCSI adapter error, adapter status = 0x20" message.
--------------------------------------------------------------------------------
Etrack Incident = ET378009
Description:
The vmoprcmd -comment <drive_index> <comment_string> command would
overwrite the serial number specified if the comment_string size was 26.
--------------------------------------------------------------------------------
Etrack Incident = ET376808 ET308657
Description:
A vulnerability has been confirmed in the NetBackup Volume Manager
daemon (vmd). By sending a specially crafted packet to the Volume Manager,
a stack overflow occurs. This is caused by improper bounds checking.
Exploitation does not require authentication, thereby allowing a remote
attacker to take over the system or disrupt the backup capabilities.
Further testing and code inspection has revealed that all other
NetBackup 5.1 daemons are potentially affected in the same manner.
Therefore, any Master Servers, Media Servers, Clients and Console machines
at this version level are subject to this vulnerability. However,
NetBackup 5.1 database agents are not affected by this issue.
Refer to the Related Document section for more details.
--------------------------------------------------------------------------------
Etrack Incident = ET366838
Description:
The robotic inventory, vmupdate, may fail after encountering a media of
different density. The problem is that vmupdate was not taking the media
type of standalone media (if available) into consideration while
determining the media type of the media to be moved into the robot.
--------------------------------------------------------------------------------
Etrack Incident = ET396055
Description:
Directories were not being listed when browsing a backup on a policy that
allowed multiple data streams.
--------------------------------------------------------------------------------
Etrack Incident = ET296412 ET326949
Description:
On Linux platforms, NetBackup now skips the backup of the following new
file system types when ALL_LOCAL_DRIVES is specified: subfs, usbfs,
sysfs, rpc_pipefs, and binfmt_misc.
--------------------------------------------------------------------------------
Etrack Incident = ET339427
Description:
Fixed the following issues with bpcatarc and bpcatrm:
- The command was failing because of an incorrect recursion level for
image Query
- The backup job submitted by bpcatarc was failing when IDIRSTRUCT was set
to 1. Now it does not submit the job when IDIRSTRUCT is 1 and it provides
a message to the user.
- When the bpcatarc output was piped to bpcatrm, it would fail to remove
the .f files that where archived in that run.
- Images in linked folders where being archived.
--------------------------------------------------------------------------------
Etrack Incident = ET387412
Description:
This pack corrects a problem where corrupt TIR fragments can be written to
tape/disk during duplication when the reader/writer of the duplication are
on the same media server.
--------------------------------------------------------------------------------
Etrack Incident = ET355632
Description:
When bpbackup is run with the option to wait for the completion of the
backup job, a firewall between bpbackup and bprd can shut down the socket
while bpbackup waits for the completion status. This causes backint to
get stuck in an infinite loop during database backups.
--------------------------------------------------------------------------------
Etrack Incident = ET395007 ET401459 ET326517 ET326956 ET329213 ET378798 ET386527 ET387430 ET399373 ET209268
Description:
Added the following robots:
Falcon Stor
HP 1/8 autoloader
Sony CSM-20
NEC T16A
Updated robot entries for the following:
ADIC Scalar DLT 448
ATL P3000
Dell PV-132T
EXB-440
Spectralogic Python
Additional Notes:
Devices can be included in this file before they are officially supported.
--------------------------------------------------------------------------------
Etrack Incident = ET303673
Description:
Added support for IBM Ultrium 3 WORM drive/media.
--------------------------------------------------------------------------------
Etrack Incident = ET400302
Description:
Non-localized error messages are displayed when deleting Volumes.
--------------------------------------------------------------------------------
Etrack Incident = ET384760
Description:
A potential race condition exists if a Windows client backup receives a
134 status (resource busy).
The race condition was caused because the bpfis process had already started
taking a snapshot and bpfis has no knowledge of the resource failure. So,
the snap is taken, but when it is completed, there is no bpbrm and bpsched
(child) process running to accept the snap information.
When the job is later re-scheduled, this (unrequested) snap information was
passed back to the new bpbrm and bpsched. The unrequested snap information
confused bpsched (child).
--------------------------------------------------------------------------------
Etrack Incident = ET349768 ET405553
Description:
"bpdbm -consistency 2" rejects images that contain files belonging to
groups with very long names, which causes media to be reused before the
retention period expired.
Workaround:
When running bpdbm -consistency record the bad images, and then import
from the media these images were on.
--------------------------------------------------------------------------------
Etrack Incident = ET320044
Description:
bpdbm's host list was not consistent with bprd's host list. Bprd
was using FORCE_RESTORE_MEDIA_SERVER correctly, however, bpdbm was
not.
Workaround:
To avoid this issue, use the original media server name, or manually edit
the image database to refer to the alternate server.
--------------------------------------------------------------------------------
Etrack Incident = ET291762
Description:
Using the Windows NetBackup Backup, Archive, and Restore interface to
Restore files and folders with curly brace characters ('{' and '}') in
their names no longer fails with the error, "- no files matched in the
given date range."
Workaround:
The workaround is to use a wildcard specification that includes these files
with curly brace characters and then delete the files that you do not want
after the restore.
--------------------------------------------------------------------------------
Etrack Incident = ET325895
Description:
Error messages were being written to the bperror log and the bpdbm log
when TIR prunning was performed. Log messages says, "stat of <backup_id>
Failed".
Workaround:
Ignore the filesfilesize value in the image file.
--------------------------------------------------------------------------------
Etrack Incident = ET401486 ET403534
Description:
If the forked re-read storage unit process detected a transient error, it
would error, then exit and remove the scheduler message queues.
--------------------------------------------------------------------------------
Etrack Incident = ET292569
Description:
RMAN - alternate client restore times out after a successful partial
restore by media server.
Though this problem was encountered while doing an Oracle alternate client
restore, the nature of the problem is such that it can happen for any
XBSA-based partial restore.
When the client is done restoring - partial restore - but there is still
data to be read, the client sockets are closed but bptm/bpdm wait for the
data to be read causing it to eventually time out with an error.
--------------------------------------------------------------------------------
Etrack Incident = ET292569
Description:
RMAN - alternate client restore times out after a successful partial
restore by media server.
Though this problem was encountered while doing an Oracle alternate client
restore, the nature of the problem is such that it can happen for any
XBSA-based partial restore.
When the client is done restoring - partial restore - but there is still
data to be read, the client sockets are closed but bptm/bpdm wait for the
data to be read causing it to eventually time out with an error.
--------------------------------------------------------------------------------
Etrack Incident = ET404676
Description:
Bprd was dumping core, and the stack trace looked similar to the following:
Core was generated by `bprd -dontfork -mpxmain'.
Program terminated with a signal 11, “Segmentation fault”.
#0 0x00034ccc in write_monitor_status ()
(gdb) where
#0 0x00034ccc in write_monitor_status ()
#1 0x0003f204 in write_clnt_monitor_status ()
#2 0x0004f7e8 in terminate_restore ()
#3 0xfdd9efc8 in __tbl_10_small_start () from /usr/lib/libc.so.1
#4 0xfdd53990 in posixgetdst () from /usr/lib/libc.so.1
#5 0x0005f9ac in run_restores ()
#6 0x00063bd4 in run_mpx_main_bprd ()
#7 0x00021ad8 in main ()
--------------------------------------------------------------------------------
Etrack Incident = ET410095
Description:
The snap driver needs to be compiled on Solaris 10 to run on Solaris 10.
--------------------------------------------------------------------------------
Etrack Incident = ET410095
Description:
The snap driver needs to be compiled on Solaris 10 to run on Solaris 10.
--------------------------------------------------------------------------------
Etrack Incident = ET351797
Description:
Specifying ALL_LOCAL_DRIVES and Shadow Copy Components caused a
bpbkar32.exe crash.
Workaround:
Do not specify any shadow copy component entries in a file list with
ALL_LOCAL_DRIVES
--------------------------------------------------------------------------------
Etrack Incident = ET410877
Description:
** If using the command line bpubsora and bpubsdb2 would core dump
immediately upon execution.
If using the jbpSA interface, and the Oracle instance was clicked on, the
application would immediately exit with the message, "bin sh:6491 abort
status 134" and it would dump a core file.
--------------------------------------------------------------------------------
Etrack Incident = ET410877
Description:
** If using the command line bpubsora and bpubsdb2 would core dump
immediately upon execution.
If using the jbpSA interface, and the Oracle instance was clicked on, the
application would immediately exit with the message, "bin sh:6491 abort
status 134" and it would dump a core file.
--------------------------------------------------------------------------------
Etrack Incident = ET315094
Description:
FBU sfr restores would not start, and bprd was slow sorting the file
history.
Workaround:
To avoid these issues, do full a restore of the FBU sfr images to an
alternate location.
--------------------------------------------------------------------------------
Etrack Incident = ET371925
Description:
Because of security relates issues, the NetBackup 5.1 MP4 patch will be
released with the 3.23.59 version of MYSQL server binary.
--------------------------------------------------------------------------------
Etrack Incident = ET409436
Description:
When backing up a control file or archived logs using an Oracle
NAS_Snapshot policy to a disk STU, the backup would fail with a
status 20 - invalid command parameter. The following is an excerpt from
the dbclient log:
10:17:31.987 [15713] <4> check_comm_file_for_error: read comm file:<10:17:20
INF - Server status = 20>
10:17:31.987 [15713] <16> check_comm_file_for_error: ERR - server exited
with status 20: invalid command parameter
Workaround:
To avoid this issue, use a tape STU.
--------------------------------------------------------------------------------
Etrack Incident = ET405301
Description:
Eliminated a two-minute delay because of changes in the Red Hat 4.0 tape
driver.
--------------------------------------------------------------------------------
Etrack Incident = ET410720 ET410648
Description:
The NetBackup Volume Manager service, or Media Manager volume daemon, logs
messages as fatal when the message is actually informational and non-fatal.
In addition, some of the messages that were logged did not contain any
information about the problem, for example:
"terminating - unknown error -- -1 (-1)"
--------------------------------------------------------------------------------
Etrack Incident = ET403523
Description:
Changes were made to extended_copy.c broke compiles for media manager
binaries.
--------------------------------------------------------------------------------
Etrack Incident = ET369929
Description:
Bpbkar debug logs are currently limited to no more than 2 Gigabytes in
size. This change enables those logs to grow beyond 2 Gigabytes.
--------------------------------------------------------------------------------
Etrack Incident = ET413509
Description:
For a policy that uses wildcards (such as, *) in the file list and enables
multiple data streams, NetBackup will not include any dot files (such as,
.rhosts) at that directory level in the backup.
--------------------------------------------------------------------------------
Etrack Incident = ET336349
Description:
The NetBackup Client for NetWare will abend on a newly installed
NetWare OES server, if it has been configured to use the Novell Boot
Loader.
Workaround:
Update SYS:SYSTEM\BPSTART.NCF to include the -B switch, which manually
specifies the server load path. For example:
-bC:\NWSERVER
--------------------------------------------------------------------------------
Etrack Incident = ET411638
Description:
NDMP DAR restore fails when a tape has been imported
or ‘FORCE_RESTORE_MEDIA_SERVER’ is set to a different NetBackup platform
(UNIX to Windows or Windows to UNIX). The restore fails with the
following error:
''NDMP_LOG_NORMAL 0 RESTORE: FILE_NAME: File does not exist in backup.''
Workaround:
To avoid this issue, use the same platform (UNIX or Windows) where the
backup was performed.
Additional Notes:
The actual problem exists on the backup site. To do DAR restore using a
different NetBackup platform (UNIX to Windows or Windows to UNIX), you
have to perform a backup using this patch.
--------------------------------------------------------------------------------
Etrack Incident = ET375812
Description:
Large volume backups and restores for FlashBackup on Windows were not
functioning correctly because of memory allocation issues and 32-bit
integer overflow conditions.
--------------------------------------------------------------------------------
Etrack Incident = ET417525 ET415832
Description:
Security permissions of some temp files were 666.
--------------------------------------------------------------------------------
Etrack Incident = ET415713
Description:
Embedded wildcard characters (when a part of a multi-byte character in some
character sets) are processed when they should not be. A specific example
of this was sjis.
--------------------------------------------------------------------------------
Etrack Incident = ET416977
Description:
Added a definition for MNTTYPE_LO so lofs is properly handled.
--------------------------------------------------------------------------------
Etrack Incident = ET418808
Description:
If a customer created an image list (bpimage -create_image_list -client
<client) after an incremental TIR backup and then ran a full or cumulative
differential backup, the TIR_INFO field in the image was not updated to 10.
As a result, TIR information was not pruned from the backup image.
Workaround:
To resolve this issue, remove image list files from the client directory.
--------------------------------------------------------------------------------
Etrack Incident = ET419597
Description:
NetBackup can get an error 130 when backing up VxFS file systems on AIX
platforms which are mounted under non-VxFS file systems that have ACLs on
them.
--------------------------------------------------------------------------------
Etrack Incident = ET419976
Description:
The disks on Hitachi arrays at higher ports cannot be backed up when using
Shadowimage, because the CSH and Port number on the disks are more than two
digits long.
The fix enables ShadowImage to backup Tagma array disks that reside on
upper ports.
--------------------------------------------------------------------------------
Etrack Incident = ET418365
Description:
Activity Monitor failure for Red Hat 4.0 (no processes listed):
For Linux, the process examination code heretofore
assumed that all files /proc/*/stat were flat files that contained process
statistics. A read() on the stat file yielded salient process information.
However, read() will fail if stat is a >directory<, whereas open() will
not fail. On Red Hat 4.0, there is at least one case where stat
is a directory and not a file (/proc/net/stat). This fix skips the
read() of stat if it is not a regular file.
Additional Notes:
This issue appears to be limited to Red Hat 4.0.
--------------------------------------------------------------------------------
Etrack Incident = ET420858 ET422804
Description:
The first line of the header of the bpclimagelist command was occasionally
not displayed. This potentially could happen for any server type, but was
seen when the master server was Linux. This would affect BMR processing.
--------------------------------------------------------------------------------
Etrack Incident = ET422477
Description:
Staging from a disk staging storage unit that contained a space would
result in images that reside on the staging storage unit to not be
duplicated properly. However, the staging job would appear to succeed in
the Activity Monitor.
Workaround:
To avoid this issue, duplicate images to a new DSSU that do not contain a
space, and change any policies that write to original DSSU to use the new
DSSU.
--------------------------------------------------------------------------------
Etrack Incident = ET324507
Description:
backuptrace would ignore bpbrm lines when the backup is multiplexed.
--------------------------------------------------------------------------------
Etrack Incident = ET423255
Description:
A change has been made to avert a potential vulnerability in a Java
authentication service that runs on VERITAS NetBackup servers and clients.
This change prohibits remote attackers from executing arbitrary code on a
targeted system. In addition, Symantec recommends that users block the
affected ports from external network access.
For more information regarding this issue, refer to the following TechNote
on the Support Web site: http://support.veritas.com/docs/279085.
--------------------------------------------------------------------------------
Etrack Incident = ET427129
Description:
The new nbdbd 3.23.59 version, was causing VISD to go down on HP-UX
platforms.
--------------------------------------------------------------------------------
Etrack Incident = ET428603
Description:
In rare cases, SCSI inquiry in page code 83 (bytes 32 and 33) returns two
digits instead of four digits of numbers.
This fix filled in the digits and enabled Shadowimage to work.
--------------------------------------------------------------------------------
Etrack Incident = ET428603
Description:
In rare cases, SCSI inquiry in page code 83 (bytes 32 and 33) returns two
digits instead of four digits of numbers.
This fix filled in the digits and enabled Shadowimage to work.
--------------------------------------------------------------------------------
Etrack Incident = ET426052
Description:
Added client support for Windows 64-bit EM64T/AMD Opteron platform.
--------------------------------------------------------------------------------
Etrack Incident = ET497068
Description:
Added Windows 64-bit client type directories so that these new types appear
as client options.
--------------------------------------------------------------------------------
Etrack Incident = ET498525
Description:
In a rare condition, one or more backups could be corrupt when using Inline
Tape Copy (ITC) with multiplexed backups. This would potentially happen
under the following conditions:
- ITC was enabled with two or more active copies.
- Multiplexing was enabled with two or more active backups.
- The schedule configuration had "continue if the copy fails" flag
set for the copy that failed in the next step.
- A copy other than the first active copy encountered "end of media", so a
new media was needed for this copy. Then, there was a problem getting
or setting up media (for example, writing the media header) such that
this copy failed before continuing the backup.
- The last buffer written to the first active copy prior to the "end of
media" on the other copy was not the first active backup.
Workaround:
To avoid this problem, do one of the following:
- Edit the schedule configuration to disable the "continue if copy fails"
flag.
- Disable multiplexing.
- Disable ITC.
--------------------------------------------------------------------------------
Etrack Incident = ET429969
Description:
Binding to a socket that was already in use caused the following error to
appear in the bpsd log file:
10061 - WSAECONNREFUSED
Workaround:
To avoid this issue, use vnetd.
--------------------------------------------------------------------------------
Etrack Incident = ET500030
Description:
bpfilter did not take into consideration, the tar header (of type LF_BAD).
Because of this, multiplex restores would go past the end of the file and
end up encrypting everything that followed the file. Ideally, LF_END and
LF_BAD would be treated alike.
--------------------------------------------------------------------------------
Etrack Incident = ET500424
Description:
VxMS version 138 did not accept certain VSS snapshot names and caused
FlashBackups to fail. Version 140 of VxMS fixed this issue.
================================================================================
=========
NB_51_3_M
=========
Etrack Incident = ET262145
Description:
Some Linux file system types (such as, tmpfs, usbdevfs, devpts, and smbfs)
were being backed up unnecessarily.
Workaround:
To avoid this issue, manually exclude the file systems that are not to be
backed up.
(NetBackup Clients: Linux2.4 Linux64 )
--------------------------------------------------------------------------------
Etrack Incident = ET295113
Description:
The scheduler was modified to correctly process a status 134 error - race
condition in a multiplex backup.
This condition occurred when a multiplex child bpsched started a backup,
but the storage unit was over committed. A status 134 error was sent back
to the multiplex child bpsched, but the child schedule was already adding
another backup to the multiplex group.
The child schedule attempted to send the new jobs information to bpbrm, but
bpbrm had exited and the connection was defunct. The scheduler was
incorrectly logging the defunct connection in the job progress log. With
this change, the scheduler will not log the broken socket, but will
correctly read and process the status 134 from the in-bound connection.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET293094
Description:
When the SUI bit is set for some admin commands on an AIX 5.2 platform,
the system OS-level command will leave a directory in /tmp. This does not
occur for the root user.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET283809
Description:
Some levels of the Linux kernel do not support netstat correctly and the
message, "Device not found," is produced.
(All NetBackup Servers: Linux)
--------------------------------------------------------------------------------
Etrack Incident = ET270622 ET270620
Description:
There are cases when a NetBackup administrator can administrate most things
on a NetBackup master server (such as policy configuration), but is not
allowed to access Host Properties on clients.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276469
Description:
When attempting to quit by using the <ESC> key in bpadm, the key is
ineffective and it fails with the error, "Attribute 254 cannot be changed".
The <ESC> key works in AIX 5.1, so this issue seems to be exclusive to the
AIX 5.2 platform.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET270167
Description:
The SCSI TEST UNIT READY timeout value has been increased from 2 seconds to
60 seconds. RS6000 media servers running many concurrent jobs to multiple
drives simultaneously could result in SCSI TEST UNIT READY error messages
that are logged into the system log.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET294713
Description:
An problem occurred that caused bpretlevel loops to echo the word, "What?,"
when changing retention levels. No input seemed to be a satisfactory
response to exit the loop until a ctrl-C was entered.
Solution:
An earlier change had changed all occurrences of gets() to fgets(), but
failed to recognize that fgets() will return the '\n', whereas gets()
did not. The subsequent strcmp was not expecting the '\n', and
always failed. The solution was to strip the [new] trailing '\n' to
emulate the gets() behavior.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET265653 ET295655
Description:
Oracle Advanced Client FlashBackup backups did not work with the new
128-bit Encryption. The failures produced an error status 25 - cannot
connect on socket. The bpfis debug log included messages similar to the
following:
17:14:27.139 [13628] <16> bpfis main: FTL - snapshot creation failed,
status 25
17:15:19.842 [13628] <4> bpfis Exit: INF - EXIT STATUS 25: cannot connect
on socket
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET278777 ET272628
Description:
Database agent backups may fail if 128/256-bit Encryption is used and vnetd
"connect-backs" to the media server is used.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET278777 ET272628
Description:
Database agent backups may fail if 128/256-bit Encryption is used and vnetd
"connect-backs" to the media server is used.
All NetBackup Windows Clients (All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET276123 ET262341
Description:
Legacy Encryption may fail if the key file on the client is large. For
example, if 39 or more pass phrases are added to a key file on a Solaris
client, encrypted backups or restores are likely to fail. The error
message "bpbrm - read() or KEYREADFD failed" will typically be reported.
Workaround:
To avoid this issue, delete some pass phrases from the key file if they
are no longer needed.
((All NetBackup UNIX Clients))
--------------------------------------------------------------------------------
Etrack Incident = ET287818
Description:
The bpimmedia -spangroups output was not displaying a span group if the
span occurred in a copy other than copy 1, and one or more preceding copies
were expired.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET287614
Description:
A Synthetic Backup Image could be expired while it was still being written.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET210298
Description:
Backups of NDMP clients were not displaying intermediate directories when
browsing the backup contents. For example, if directory A/B/C was backed
up, /A and /A/B were not appearing when browsing the backup.
Although the NetBackup catalog contained the directory information, it did
not return it when the directory was not explicitly backed up. This fix
will enable browsing of these NDMP backups even if they were created prior
to the fix.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET297729
Description:
Oracle BLI backups failed on Solaris 10 when trying to backup multiple VxFS
partitions. The following would be logged:
For second or later file system, checkpoint mount may fail, bpbkar log:
11:58:55.199 [8353]<4> bpbkar set_clone_context: /usr/sbin/mount -F vxfs -o
ckpt=NetBackup_full_orac101_1106243854
/dev/dsk/c1t0d0s4:NetBackup_full_orac101_1106243854
/tmp/NetBackup_checkpoints/NetBackup_full_orac101_1106243854 >/dev/null
2>/dev/null
11:58:55.464 [8353] <16> bpbkar set_clone_context: ERR - Cannot mount
checkpoint NetBackup_full_orac101_1106243854 on
/tmp/NetBackup_checkpoints/NetBackup_full_orac101_1106243854.
(NetBackup Clients: HP HPIA64 RS6000_433 Solaris Solaris9 )
--------------------------------------------------------------------------------
Etrack Incident = ET296425
Description:
This bpbrm binary fixes an intermittent socket problem that occurred
during database backups. The backup would fail with a status of 54.
An examination of the bpbrm debug log showed the following type of message:
07:10:52.662 [3832.2768] <16> bpbrm listen_for_client: listen for client
timeout during accept from name listen socket after 60 seconds
An examination of the dbclient debug log showed the following type of
message:
02:24:07.956 [528.660] <8> connectSock: WRN - connect() to server failed,
No connection could be made because the target machine actively refused it.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET299437
Description:
An extended value (90 days, for example) for the "Keep Logs for x Days"
global configuration setting appears not to work. The debug logs are
still deleted after about a month on NetBackup media servers that
are included as part of the catalog backup configuration.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET272516
Description:
NetBackup sometimes incorrectly handles an exclude list with an asterisk
character "*" as an entry and an include list.
Workaround:
To avoid this issue, use either "/" or "/*" in the exclude list rather
than "*".
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET294004
Description:
Backing up files which do not have a group name can fail with a Status 13.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET291947
Description:
A SAP online backup is started and a Tablespace (PSAPES46CI) in the Oracle
database is placed in BACKUP MODE by the agent. However the backup receives
a STATUS 134 error in BPTM & BPBRM. The bpbkar, bptm, and bpbrm processes
exit and clean up after themselves. The SAP agent process remains active,
awaiting the backup to be queued and run again, leaving the Tablespace
(PSAPES46CI) still in BACKUP MODE.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET294365
Description:
vmd would hang after 21 days under VxSS/NBAC.
(All NetBackup Servers: Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET277025
Description:
When a local file system was auto-mounted to itself, NetBackup would
incorrectly skip the backup of the local file system on SGI systems. This
occurred when the hostname returned from the uname command and the
hostname stored in /etc/mtab were not both fully-qualified or both short
(for example, uname returns "machine" and /etc/mtab contains
"machine.domain.com").
Workaround:
Make sure the hostname returned by the uname command and the hostname
specified in /etc/mtab both use either the fully-qualified hostname or the
short name.
(NetBackup Clients: SGI65 )
--------------------------------------------------------------------------------
Etrack Incident = ET300673
Description:
If a /dev/rdsk device was used as the extended copy command target, the
code to convert the disk path to the sg path for scsi-passthrough commands
fails if the disk is managed by the ssd driver. This caused an error to be
logged in the bptm log this was similar to the following:
07:00:44.807 [13213] <2> setup_mover_tpc: open of pass-through path
/dev/rdsk/c1t50060E80034FED11d0s2 failed, unable to open
/devices/pci@1f,0/pci@5/SUNW,qlc@1/fp@0,0/sgd@w50060e80034fed11,0:raw, No
such file or directory
Workaround:
To avoid this issue, configure the corresponding sg path directly in
mover.conf.
(All NetBackup Servers: Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET299670 ET300174 ET300175 ET300176
Description:
bpbrm would hang if a database agent backup failed in a certain way.
The job would appear to be running in the Activity Monitor until an
administrator terminated it.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET301501
Description:
A Scheduler race condition was occurring: Child sched was sending a
‘done’ to the main sched while the main sched was sending a ‘term one’ to the
child sched. The child sched erroneously treated the ‘term one’ as the
response to the ‘done’ and then exited prematurely.
The correct child response is to discard (ignore) the ‘term one’
request and wait for the main sched response to the 'done' command.
Also missing was code to discard (ignore) a ‘suspend’ request after having
sent a 'done' to the main sched.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET300757
Description:
Interval-mode backup windows open longer than 6-days (144-hrs) resulted in
continuous backups on the seventh day, till the window closes.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET277470
Description:
Creating a long name in the hosts file and using that in a policy causes
various parts of BEDS to crash.
Workaround:
Ensure all names used to reference the host are 15 or fewer characters in
length
All NetBackup Windows Clients
--------------------------------------------------------------------------------
Etrack Incident = ET299481
Description:
ltid on Linux, when starting up vmd, causes the following to be logged
in /var/log/messages:
kernel: application bug: vmd(16703) has SIGCHLD set to SIG_IGN but calls
wait().
kernel: (see the NOTES section of 'man 2 wait'). Workaround activated.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET300679
Description:
Restoring compressed files from a FlashBackup on a Windows image resulted
in tar32 aborting when the master server was not a Windows platform.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET299843
Description:
Bptm, bpdm, or bpdbm would occasionally crash during TIR incremental
backups. Bprd, bpbrm, or bpsched were possibly affected as well.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET302767
Description:
The NetBackup Windows Administration Console will crash when attempting to
view certain jobs details in the Activity Monitor.
(All NetBackup Servers: Intel-NT Intel64)
--------------------------------------------------------------------------------
Etrack Incident = ET296072
Description:
When installing NetBackup Server on SUSE 9 systems, the installation
appeared to hang.
Additional Notes:
When the NetBackup 5.1GA install hangs, the make_scsi_dev process must be
killed from another terminal session. The NetBackup install should be
allowed to run to completion after make_scsi_dev has been killed. The
NetBackup 5.1MP3 Maintenance Pack must then be applied, and the
make_scsi_dev command run. Once make_scsi_dev completes, devices may be
configured.
(All NetBackup Servers: Linux)
--------------------------------------------------------------------------------
Etrack Incident = ET299846
Description:
Multiplexed backup jobs would occasionally hang. This would happen if the
bptm or bpdm parent process crashed, or if bptm could not successfully add
the TIR fragment for a TIR backup. It may also happen because of a timing
problem in bpbrm when new jobs are started.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET292021
Description:
A new bperror -jobgroupid option is introduced to search the error database
based on the job group ID. This is helpful for an Inline Tape Copy job
because it may log to multiple job IDs.
The syntax of the option is:
-jobgroupid job_group_id
where job_group_id is the job group ID number for the job.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET303111 ET301828 ET275764 ET293905 ET279264 ET284804 ET285117
ET270133 ET295671 ET295695
Description:
Added support for the following libraries:
EXABYTE VXA-2 1x10 Autoloader
EXABYTE 1x7 Autoloader
Added support for the following drives:
EXABYTE VXA-2
HP VS160
STK VDRIVE
Updated the following libraries:
COMPAQ TL895
DEC TL810
DEC TL895
HP MSL6000
HP MSL5000
HP UHDL
IBM 3583
IBM 4560SLX
IBM 7337
SONY PETASITE CSM200/100/60
Additional Notes:
Devices can be included in this file before they officially supported.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET293146
Description:
NetBackup processes may core if an error occurs. The stack from the core
file showed the processes were in the ndmp_user_scsi() or ndmp_user_tape()
functions.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET286656
Description:
bpverify would fail with a status 191 on a single master/media server
because of a socket timeout.
Workaround:
To avoid this issue, do not run bpverify on very large backups with a slow
or stressed network.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET266342
Description:
bplist returned duplicate listings for a DB2 file; and the .f file showed
only one instance of the file. This problem did not occur for other similar
client backups. If IMAGE_LIST file was removed, the file would be listed
only once.
Workaround:
To avoid this issue, turn off image indexing.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET303564 ET302345
Description:
The main sched failed to recognize a media server that was initially
off line, and then brought on line after the main sched went active.
If the media server was brought back while the scheduler was not active or
the master server was bounced, then everything worked correctly.
This fix enables the scheduler to ‘pick up’ the newly introduced media
server by using the ‘/bp/bin/admincmd/bpschedreq -read_stunits’ command.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET302345 ET303564
Description:
Backups using in line tape copy (ITC) and storage unit groups, failed to
give preference and select the local storage unit.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET303449
Description:
Backup failed with disk full even though there was still room on the disk.
Workaround:
The problem is the result of the settings in /etc/security/limits. The
value for file size could be increased to eliminate this problem.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET273869
Description:
CLUSTER_NAME is behaving the same way as if REQUIRED_INTERFACE was set.
Additional Notes:
This fix only applies to UNIX platforms. For Windows, the behavior has
always been as described. To turn off this behavior, add the following entry
in bp.conf:
ANY_CLUSTER_INTERFACE = 1
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET299846
Description:
bpdbm crashed due to a problem in a CPR case.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET304308
Description:
Staging/Duplication from a disk staging unit was not working when more
than 32765 files were in the staging area.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET270494
Description:
AIX would report the swap size in bytes and not kilobytes.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET309223
Description:
Backups were taking too long between job complete and 'end writing'.
To fix this problem, a correction was made to a recursive call in
searching for a storage unit.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET289123
Description:
Staging of images which were no longer in the catalog caused a 190 error.
Workaround:
Manually delete images which are not in the catalog.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET287633
Description:
When simultaneous, differential, incremental backups are done for the same
client, a subsequent synthetic backup for that client may fail.
Workaround:
To avoid this issue, do not run simultaneous backups on the same client.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET309223
Description:
Backups were taking too long between job complete and 'end writing'.
This problem was fixed by correcting a recursive call in searching for a
storage unit. In addition, removed a media server version check in the
main ITC - find storage unit loop.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET309085
Description:
On Solaris 10 platforms, NetBackup now skips the backup of the new file
system types ctfs and objfs.
Workaround:
Add the mount points of any objfs and ctfs file systems into the exclude
list on that client. For example, /system/object and /system/contract.
(NetBackup Clients: Solaris9 )
--------------------------------------------------------------------------------
Etrack Incident = ET321994 ET266342
Description:
While testing the NetBackup 5.1 MP3 patch, the ability to browse ASCII
catalogs did not exist.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET323188
Description:
Disk-staging of a directory with no images to stage causes the program to
terminate without informing its parent program that it has finished.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET323222
Description:
Disk-staging only stages the one copy of an image in the directory.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET320557
Description:
When a resume backup exits, it would attempt to send a request queue
message to the GUI. But the GUI does not read these messages, so the
message would stay queued until the main sched exited.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET323183
Description:
If the stream in control snap failed on a WOFB, then the other streams were
allowed to continue a non-snap backup. This was the correct operation. The
problem was that the non-control streams performed a backup of all files in
the file list, but they should have only backed up their specific files.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET316635
Description:
User backups failed to check if an exiting backup was the last backup job
from a specific SPID before removing all user backups from the work list.
This oversight caused the removal of active jobs from the work list, and
resulted in a sched core.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET323827 ET324060
Description:
DSSU stage to tape fails with a status 134 if the media becomes full.
A NON_MPX_RESTORE can cause bptm to select the wrong media during a restore
operation that spans tape.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET303490 ET315602 ET324213 ET323143 ET323736 ET277364
ET315274
Description:
Added the following library:
CERTANCE LTO3 CLL 6400
Updated the following library entries:
ADIC 10k
ADIC Scalar 24
ATL M2500
ATL M1500
Exabyte 480
IBM 3583
Additional Notes:
Devices can be included in this file before they are officially supported.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET315056
Description:
WORM tape support uses sixteen byte SCSI-3 commands. Some HBA's have
problems with these commands, and can cause SCSI Bus Reset's. A new touch
file, volmgr/database/NO_SIXTEEN_BYTE_CDB, can be created to disable these
commands. This will effect your ability to use WORM tape with some drives.
Solaris platforms have been found to not handle sixteen byte commands on
Parallel SCSI HBA's, so devices on these HBA's are automatically prevented
from using sixteen byte SCSI commands.
Workaround:
To avoid this issue, use an HBA that supports sixteen byte commands.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET384887
** Description:
ATTENTION: A related TechAlert exists for this issue.
NetBackup 5.1 Maintenance Pack 3 contained a condition under which data
loss might have occurred. This data loss would occur if the following
conditions were true:
NOTE: This issue only affects duplicate images using NetBackup 5.1_MP3.
The original or "source" image remains valid.
1. The original backup image must be written to tape. Virtual Tape
Libraries (VTL) will be affected by this as well. Disk-based backups are
not affected by this issue.
2. The original backup image must span more than one fragment. This would
be caused when multiplexing is enabled, when a backup image spans more than
one tape, when the maximum fragment size is reached, or when a checkpoint
occurs for checkpoint-restart backups.
3. The image must be a duplicate image and meet one of the following
conditions:
- The source image was not multiplexed on tape.
- The duplication is performed with the "Preserve Multiplexing" option
disabled.
If all the criteria above apply, then the second and subsequent copies of
backup images created by the duplicate operation will be written incorrectly.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
=========
NB_51_2_M
=========
Etrack Incident = ET218330
Description:
The synthetic backup timing was incorrect because to the time it took to
read the large extents.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET257821 ET232038
Description:
A large number of down storage units would result in and apparent scheduler
hang. In reality, the scheduler was timing out to each media server in
series. This fix recognizes and skips duplicate attempts to a down media
server.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET220801
Description:
"Ghost" backup jobs were appearing in the activity monitor. For example, a
deleted backup job could (partially) re-appear in the activity monitor.
This would only happen to a deleted job that was one number lower than a
multi-copy backup.
And example of this scenario is:
1) Job n finishes.
2) Job n is deleted from the activity monitor.
3) Multi-copy job n+1, n+2, n+3 finishes.
4) Activity monitor incorrectly updates jobs n, n+1, n+2.
5) Deleted job n (partially) re-appears.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET212797
Description:
The interrupt code paths for the Scheduler were simplified to eliminate
scheduler core dumps because the interrupt handlers were too "noisy".
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET211907 ET210846
Description:
If an in-line copy backup is multiplexed, and one of the copies gets a
failure (but continues), then subsequent additions to this troubled
multiplexed group generates excessive failure messages.
This code change inhibits adding new backup jobs to multiplexed groups
that contain a failing copy.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET218162
Description:
The bpplschedrec -deldayomonth command deletes a range of days instead of
the specified day.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET230710
Description:
A warning message is produced in RedHat 9 that indicates NetBackup is
issuing a wait after telling the operating system (OS) to ignore the
relevant signal. The OS disregards the wait so there is no actual
problem. This only affects bpcd in standalone mode.
Workaround:
The current solution for this issue is to ignore the warning message.
All NetBackup Windows Clients (All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET236426 ET230514
Description:
If NetBackup is configured to duplicate an image and also configured to
change the retention level of the new copy to a retention level that is
configured as "expires immediately", then the original image also appears
to be expired at the end of the duplication. Any backup taken after that,
returns the original image.
Workaround:
Any subsequent backup would expunge expired copy from the image file, and
bring back the original image (to image list and for restore operation)
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET234522
Description:
SCSI-3 library commands have standardized on op code 0x37 for the command
INITIALIZE ELEMENT STATUS w/RANGE. Previous implementations by numerous
vendors had used vendor-unique code 0xE7 - which was implemented by
NetBackup. SCSI-3 libraries will return sense data 20/00 when sent an E7.
Workaround:
To avoid this issue, use only tapes with barcodes that are valid for your
library vendor.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET234298
Description:
NetBackup was not properly using the firmware revision to retrieve the
correct device mapping in the device mappings file. This has never been a
problem in the field because VERITAS has never had to use the firmware
revision in the device mappings file - this is being done for new device
support.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET235646
Description:
Previously, cluster_config -a would go into an infinite loop when
configuring add-on products for monitoring when anything but a lowercase
"y" was entered at the following prompt:
You have chosen to monitor <addon>. Is this list correct? [y,n] (Y)
This issue has been resolved and the defaults, "Y" and "y" are all valid
responses.
Workaround:
To avoid this issue, you would have entered a lowercase "y" instead of an
uppercase "Y", or you could have taken the default.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET223542
Description:
The scheduler does not automatically start "bpexpdate -deassignempty" to
deassign media that do not have associated images after "bprecover -l" has
been run.
Workaround:
If a bprecover -l has been executed, remove the following file:
UNIX:
/usr/openv/netbackup/bin/deassign_lock
Windows:
VERITAS\NetBackup\Bin\deassign_lock
Do not remove this file if a catalog recover (bprecover -r) has been done.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET260510 ET259248
Description:
Only fails on a UNIX master. If a user archive fails to remove some files
(and generates a status 4), then the .f file is removed, and the files, up
to the failed file, are removed from the client.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET235410
Description:
It was possible that the bpjobd process may hang under certain conditions.
It has been demonstrated that this can happen if bpduplicate and bptm
processes are killed, but it may occur in other situations. The bpjobd
process does not die, it just loops consuming CPU time, and nothing is
written to the debug log. This prevented the job DB from being updated
with the status of existing jobs, or new jobs. bpjobd must be killed and
restarted by the system administrator to resolve this problem.
This problem is indicated by the following data in the bpjobd debug log,
referring to non-existent job ID 3:
18:45:00.734 [17433] <2> process_active_job: Frame type is 3 (jobid 3)
Sock (93)
18:45:00.734 [17433] <2> process_active_job: Adding provider(1) for job(3)
socket (93)
18:45:00.734 [17433] <2> process_active_job: Begin to write (JOBTRYFILE) at
offset 68 job (3)
Workaround:
To avoid this issue, kill and restart the hung bpjobd.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET223801
Description:
Under some circumstances, an image file might be expired on the Master
server and remain on the Media server. When media full is detected for a
DiskStaging storage unit, the media server attempts to remove the two
oldest images. If these images were no longer in the catalog on the Master
server, they could not be removed and the backup would fail.
The code was changed to remove the image files if the Master server no
longer had them in the catalog.
Workaround:
To avoid this issue, manually delete the image files that are causing the
problems along with their .ds file.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET221591 ET185355
Description:
The predict algorithm was not sorting correctly. It failed to position
cumulative incremental backups before the differential incremental backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET232060
Description:
To assist image database cleanup in 24x7 environments, image database
cleanup will always run in background. bpsched will not wait until the
cleanup is complete before scheduling new jobs. In addition, bpsched will
initiate an image database cleanup at periodic intervals, 12 hours by
default. The time interval may be changed by using the
bpconfig -cleanup_int <hours> command. For example, to change the interval
to 24 hours, execute the following command:
bpconfig -cleanup_int 24
Use a value of 0 to disable bpsched from starting the periodic image
database cleanup. The maximum value is 744 hours. Bpsched will still
perform an image database cleanup at the end of a session, however, it will
be run in background. On UNIX systems, bpadm may also be used to change
the interval.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET231652 ET224755
Description:
NetBackup was not properly examining the SCSI library mail slot status
bits; therefore, during ejects of more than a MAP-full of media, the eject
operation may abort after the first MAP-full is removed.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET257819
Description:
ltid will now terminate gracefully on UNIX platforms if it detects a
(invalid) device configuration where non-shared robots (robots with types
other than TLD, TL8, TLH and LMF) are configured on media servers with no
drives configured.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET232896
Description:
A problem exited causing a snapmirror-to-tape backup to fail with an
error 99.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET233382
Description:
In one instance, some catalog files were corrupted when the system ran out
of disk space. This caused bpdbm to crash whenever it tried to load the
corrupted catalog files. The command "bpdbm -consistency 2" also crashed
every time these corrupted files were checked.
The cause of this problem was the buffer overrun in libcatalog when loading
corrupted files: the pointers and string lengths are wrong after mapping
buffer overrun to fix the crash of bpdbm. In addition, code was added in
image.c to move corrupted catalog files out of /db/images (into /db.corrupt
directory) when running "bpdbm -consistency 2".
This new code resulted in the following new feature:
"bpdbm -consistency" will move corrupted catalog files out of the catalog
repository "/netbackup/db/images" and into "/netbackup/db.corrupt"
directory.
After running the above command, users will need to do the following:
1. Record all file paths in /db.corrupt.
2. Re-import the files in step 1.
3. Delete files in /db.corrupt.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET225198
Description:
The Phase II import of (non-binary) catalog files can fail because the
code tries to re-name an imported file before closing the file.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET221883
Description:
The file list for stream 1 is listed twice in the job details when stream 1
is force-started to take a snapshot. This error is only on multi-stream
snapshot backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET223787
Description:
The scheduler debug log was filling with the following message:
"storage unit was not assigned for clientjob <host name> <policy name>
<stunit name> <copy>"
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET260902 ET259243
Description:
When a user imports and/or restores a NetBackup 3.4 backup image in
NetBackup 5.0/5.1, files greater than 1GB in size fail to restore and
produce, "disk full at byte xxxxx", errors in the restore log. Image verify
also fails with a gigabytes size mismatch error for those files. In
addition, the Backup, Archive, and Restore GUI shows a discrepancy in
reporting the size of those files.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET258421 ET258293 ET261720 ET221167
Description:
When a database agent gets queued again, multiple times in a very short
period of time, there were cases where the server would time out waiting for
the client to connect. This fix prevents the client from getting behind the
server in these instances.
(NetBackup Clients: ALPHA HP Intel-NT Intel64 Linux2.4 RS6000_433
RS6000_51 Solaris )
--------------------------------------------------------------------------------
Etrack Incident = ET223261
Description:
Bpbrm can core dump if one of the signal handlers interrupts a malloc
or free when the code is updating the pointers that are used to track
the free or allocated space.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET264088
Description:
A multiplexed TIR duplication would fail if the prior backup duplication
had a TIR size greater than 1 Gigabyte, and would lead to a bptm crash.
Workaround:
Retry the duplication using non-multiplexed duplication. It is also
possible that the duplication will succeed by trying multiplexed
duplication.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET262983 ET262880
Description:
The backup job start time in the Activity Monitor was the same value for
each attempt. The start time value should be different on each attempt.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET192195
Description:
The ltid daemon will report operational issues at startup time using
localized error messages. In some cases, the boot-time console device can
not properly display non-ASCII characters, and can cause the boot-time
error conditions to be easily understood.
All of the error messages that can be presented at startup time are now
printed in ASCII only. No localization of the error messages will be
attempted by ltid.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET258458 ET167719
Description:
Added support for the following drive and media types for ACS:
Drives:
HP LTO-3
Media:
LTO-400G
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET263758
Description:
The scheduler would send a message on the wrong queue when it detected an
error on a client snapshot. This may result in a scheduler crash, if the
error happened frequently, or if the scheduler stayed running for a
prolonged period of time.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET265188
Description:
If an SSO drive is DOWN on all servers, and its scan host unregisters
(services are shutdown), vmd/DA will not transfer this drive to another
scan host. If the drive is brought UP on a server(s), it will have no
scan host, and will stay in the NO-SCAN state.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET265663
Description:
The image resume time limit was not always being updated on Checkpoint
backup jobs. Multiple attempts to resume a job might result in a backup
that is restarted instead of resumed.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET229709
Description:
The following bptm logging statement was confusing and has been changed.
09:37:46.035 [16536] <4> ndmp_internal_open_and_auth: NDMP server DEBUG
suggests protocol version 3
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET262921
Description:
The error "license use has been exceeded", status code 159, may have been
erroneously generated when using a SAN media server license for a media
server that is part of a cluster.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET212904
Description:
A failure occurred when attempting to restore from an image that was
backed up with the following setting:
"SET FILES = ( Multiple Files ) "
A status 5 would be generated.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET266252
Description:
There are two issues that may have caused backups to fail:
1. bpsched was not updating the RESUME_EXPIRATION time on failed backups.
2. bpdbm did not update the REQUEST_PID on subsequent Q_IMAGE_NEW requests.
Workaround:
To avoid this issue, make sure any job that has been resumed is finished
before the cleaning is done.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET233105
Description:
vmphyinv is unable to read the tape header in NDMP drives. This affects
NDMP drives and the NetWare media server (that are handled using NDMP
protocol).
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET233201
Description:
Some FlashBackup backups took longer on NetBackup 5.0 than they did for
NetBackup 3.4 because of the overhead of the binary catalog.
Workaround:
To avoid this issue, raise the maximum files-per-add value using the
/usr/openv/netbackup/MAX_FILES_PER_ADD to 9500. This will raise the limit
for all backups.
Additional Notes:
The number of files added per call to the catalog has been raised to 9500
from 500 for FlashBackup. This limit can be modified by putting the desired
value in /usr/openv/netbackup/FBU_MAX_FILES_PER_ADD, which only affects
FlashBackup backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET267530
Description:
Scheduler code was too inflexible and did not allow for easy addition of
new Windows Open File Backup (WOFB) snapshot clients. This change corrected
the scheduler and eased the addition of future Microsoft clients.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET263530
Description:
The method of configuring inetd has changed in Solaris10. Changed scripts
that deal with the modification of inetd.conf to use the new method if on a
Solaris 10 machine (Sparc or X86).
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET263530
Description:
The method of configuring inetd has changed in Solaris 10. Changed scripts
that deal with the modification of inetd.conf to use the new method if on a
Solaris 10 machine (Sparc or X86).
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET233198
Description:
bpSALinfo needs to be linked with a more recent set of SAL API libraries to
be able to interface to ccstor 4.0, and SPC 3.4. Minor changes are needed
to account for differences in return values for ccstor 4.0.
(All NetBackup Servers: HP RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET268568
Description:
The scheduler would take a long time scheduling multiple data stream jobs
if there were many policies that were due to be auto discovered at once.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET232056
Description:
When selecting a media ID for backup, if the allocation date in the media
database does not match the assigned date in the media manager database,
FREEZE the media because there could be database inconsistencies.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET268528
Description:
The cluster monitor script would not restart the stopped daemons on some
platforms. This happened because the .NetBackup_STATE file did not have a
NEWLINE at the end of the file, so SED did not see it as a complete line
and did not return a needed result.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET266303
Description:
For NDMP restores on NetApp filers, more than 1024 files can be restored
using DAR.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET269452
Description:
This version of the SG driver adds support for Solaris 10.
(All NetBackup Servers: Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET263450
Description:
Errors occurred when the Sony TSL-11000 tape library would put the
cleaning tape away to its appropriate slot.
Workaround:
The cleaning actual works, but the cleaning statistics do not get updated.
Running a tpclean -M <drive name> after the failed clean will correct this.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET269768
Description:
Add support for RedHat 3.0 on the Zseries platform. This required a new
client type which is called IBMzSeriesLinux2.4.21. The number is based off
the operating system (OS) kernel.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET269768
Description:
Add support for RedHat 3.0 on the Zseries platform. This required a new
client type which is called IBMzSeriesLinux2.4.21. The number is based off
the operating system (OS) kernel.
(NetBackup Clients: new )
--------------------------------------------------------------------------------
Etrack Incident = ET233695
Description:
During multiplexed restores, when multiple tar instances are restoring
files onto the same directory tree, some of the files may not get restored
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET264130 ET266195 ET265296 ET262877 ET258096 ET223049
ET258202 ET258252 ET257553 ET230175 ET223889 ET223643
Description:
Added support for the following libraries:
- ATL M2500 (fixed inventory issue)
- Certance CLL3200 (fixed TLD(AVR) issue)
- HP DAT72X6
- IBM 3623 2SX
- IBM DDS4
- NEC T40A
- Sony Lib-162
- Sony Petasite CSM200/100/60 (serialization change with 4.01)
Added support for the following drives:
- Benchmark VS640
- Sony SDX-900V
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET269914
Description:
BLI failed to mount a clustered storage checkpoint. The bpbkar log
displayed the following:
14:56:41.914 [13792] <4> bpbkar set_clone_context: /usr/sbin/mount -F
vxfs -ockpt=NetBackup_full_test_1096401397 /dev/vx/dsk/cfs_dg/rb_test
NetBackup_full_test_1096401397 /rb_test_s/NetBackup_checkpoints
NetBackup_full_test_1096401397 >/dev/null 2>/dev/null
14:56:42.409 [13792] <16> bpbkar set_clone_context: ERR - Cannot mount
checkpoint NetBackup_full_test_1096401397
on /rb_test_s/NetBackup_checkpoints/NetBackup_full_test_1096401397.
(NetBackup Clients: HP RS6000_433 Solaris )
--------------------------------------------------------------------------------
Etrack Incident = ET269541
Description:
An extra invocation of a function was resetting the delay time on tape UN
mounts for user backups. Removal of the extra invocation should produce
more optimal UN mounting of tapes when performing user backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET269768
Description:
Added support for IBMzSeriesLinux2.4.21.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET235418
Description:
In certain configurations, calls to gethostbyname (3nsl) may fail. In such
cases, VERITAS did not force a DNS lookup for a fully qualified host name.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET270921
Description:
Because of a race condition, ACSLS drives may get downed during mount
processing.
Additional Notes:
The signature of this race condition is that the event viewer (or syslog)
will note errors in semaphore operations from ACSD, and then take the
drive down.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET274324
Description:
When multiple disk staging jobs are initiated and the DSSU is at or
near capacity, some jobs may fail with a status 84 error, "disk full,
attempting to remove the same staged images".
BPDM logs contain entries that indicate no matching entries found in the
catalog, (db_IMAGE: db_end_sts() failed: no entity was found).
Workaround:
If the DSSU is nearing capacity, manually remove the previously staged
images. These images are identified by an additional zero-length file of
the same name, with a ".ds" extension.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET274137
Description:
A bug in write_data_tir() had corrupted the TIR file when the disk was full
and when writing to a disk staging unit.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276040
Description:
Maintenance Pack 1 for the NetBackup 5.1 release introduced a performance
issue with NDMP DAR restores for catalogs with a large number of files.
This fix corrects the performance problem.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276089
Description:
Added 64-bit Oracle support for Linux on the AMD64 and EM64T platforms.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276089
Description:
Add 64-bit Oracle support for Linux on the AMD64 and EM64T platforms.
(NetBackup Clients: Linux2.4 )
--------------------------------------------------------------------------------
Etrack Incident = ET272137 ET274040
Description:
Corrected a case where incremental backups were backing up complete data
sets (full backup). This happened only for multi-stream WOFB or advanced
client where a previous non-control stream backup failed.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET273464
Description:
Corrected a problem where tapes would not unmount in a reasonable amount of
time.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET271482
Description:
Re-naming a schedule, and then attempting an immediate backup using the
newly named schedule caused a core dump in the scheduler.
Workaround:
To avoid this issue, do not add, remove, or modify policy or schedule
information and then attempt an immediate backup while scheduler is
running.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET275271
Description:
In a rare circumstance, the last block of a multiplexed, duplicated copy of
a TIR backup may be corrupt. A verification of the copy would not be
successful. The original backup is not affected. This does not affect
non-multiplexed duplications of a multiplexed TIR backup.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276969
Description:
Corrected a problem that allowed an incorrect block size to be put in tape
backup headers. This problem occurred when the file SIZE_DATA_BUFFERS
exists to change default tape block size, a backup is duplicated to tape
and the source/destination media is on the same machine. The incorrect
block size in the tape backup headers would not prevent restores from
working, however, it would prevent import phase 2 from working if that
copy (with the incorrect backup header) was ever imported.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET276830
Description:
User archive or synthetic backups that have checkpoint enabled, and receive
a 134 (resource busy) will be rescheduled. But the Activity Monitor lists
the job as "done" status 134.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET272608
Description:
If a set of configuration options in bp.conf is very large, it can cause
bpgetconfig and the GUI to not display any of the items. This applies to
items in bp.conf with the same name, for example CONNECT_OPTIONS or SERVER.
Additionally, if that set of configuration options is edited in the GUI,
for example, to add a new server to the CONNECT_OPTIONS, all of the
previous data for that set of options will be lost and only the new data
will be in bp.conf.
Workaround:
To avoid the issue, reduce the number of configuration options with the
same name.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET273254
Description:
NDMP restores fail with Hitachi NAS when restoring the entire backup image.
The NDMP log shows the path being sent twice in the NDMP_DATA_START_RECOVER
message.
Workaround:
Restores will work as long as the file system path is not selected (the
path that was initially backed up). If the entire image is needed for
restore, select all the files and directories in the file system path
instead.
(All NetBackup Servers: HP Intel-NT Linux RS6000 Solaris)
--------------------------------------------------------------------------------
Etrack Incident = ET272608
Description:
If a set of configuration options in bp.conf is very large, it can cause
bpgetconfig and the GUI to not display any of the items. This applies to
items in bp.conf with the same name, for example CONNECT_OPTIONS or SERVER.
Additionally, if that set of configuration options is edited in the GUI,
for example, to add a new server to the CONNECT_OPTIONS, all of the
previous data for that set of options will be lost and only the new data
will be in bp.conf.
Workaround:
To avoid this issue, reduce the number of configuration options with the
same name.
All NetBackup Windows Clients All NetBackup PC Clients
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET258096 ET270045 ET268600 ET260459 ET273679 ET220083 ET260459
ET274370
Description:
Added support for the following libraries:
- ARCHIVE Python 06241-XXX910B
- Certance CLL3200
- Diligent VTF Open
- IBM DDS4 Autoloader
- IBM 3623 2SX
- SEAGATE DAT 9SP40-400910B"
Added support for the following drives:
- HP LTO3
- IBM Ultirum-TD3, 3580-TD3
Additional Notes:
Devices can be included in this file before they officially supported.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET259704
Description:
When running Windows Open File Backup enabled backups using VSS on
Windows 2003 clients, performance of the backup is considerably slower
compared to when running a Windows Open File Backup enabled backup using
VSP or with Windows Open File Backup disabled. Only backup performance is
affected and no data loss occurs as a result of using VSS as the Windows
Open File Backup snapshot provider.
All NetBackup Windows Clients
--------------------------------------------------------------------------------
Etrack Incident = ET230111
Description:
When backing up encrypted data on Windows 2000 clients using Windows Open
File Backup to backup an open/active file, the encrypted data will be
inaccessible to the backup, and the backup will end partially successful.
Additional Notes:
In NetBackup 5.1 MP2, Windows Open File Backup snapshots will be disabled
when backing up encrypted data on Windows 2000 clients. This is not
applicable to Windows operating systems newer than Windows 2000 such as
Windows XP and Windows 2003. A logging message similar to the one below may
appear in the bpbkar32 debug logs indicating this on Windows 2000 clients:
4:29:09.328 PM: [2620.2648] < 4> file_access::V_OpenForRead: INF -
Disabling
snapshots for: D:\lock\ORW_SN.txt
By disabling snapshots for Windows encrypted data on Windows 2000 clients,
these files may not be accessible to the backup if they are busy or locked
by another process. If Windows Open File Backup protection is important,
Windows encryption should be disabled on the files being backed up.
All NetBackup Windows Clients
--------------------------------------------------------------------------------
Etrack Incident = ET271139
Description:
Image fragments are sometimes left behind on the disk storage unit after
the image expiration has occurred, and the accumulation of such "rogue"
fragments can eventually create backup failures unless the rogue fragments
are manually deleted by an administrator.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET279250
Description:
Backups of encrypted files completed with a "partially successful"
message, however, the encrypted files were not backed up.
(NetBackup Clients: Intel-NT )
--------------------------------------------------------------------------------
Etrack Incident = ET283490
Description:
Initiating multiple Oracle backups caused crashes in scheduler. Each of the
Oracle backups attempted to read/update/write a group ID file (used for
synching snap backups), but there was no lock mechanism on the file access,
so each Oracle backup had the potential to corrupt another’s file access.
When this problem was discovered, the environment was running about
20-Oracle backups (one policy per client).
Workaround:
To avoid the issue, stagger window openings for Oracle backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET279063
Description:
When performing duplication and direct shared memory was being used instead
of sockets, increase the timeout the writing bptm/bpdm waits for the
reader to setup the shared memory. Previously, if the system was
very busy, the timeout was not enough.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET277797
Description:
Robot inventory update was not taking into account the media type for
existing standalone media that have been moved into a robot. It was only
looking at the barcode rules. This was the basic cause of the error "media
type and volume group mismatch (101)". Robot inventory update now updates
the media residences to be robotic without changing the media type.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET284896
Description:
When running a Restore job, the information about the job does not appear
in the Job Details -> Job Overview tab in the Activity Monitor, however,
the job still completes successfully.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET287099
Description:
After running FlashBackup backups that were aborted, a snapcachelist
command showed the snapshot cache device still in use even though the
snaplist command did not return any snapshots. Eventually, the snapshots
would fail because of the 32 snapshots per cache device. A reboot is
required to recover the use of the cache device.
Additional Notes:
The snap driver has been modified to use a different flag bit to denote a
snapshot in creation versus a snapshot that is being deleted. A snapshot in
creation is allowed to be listed by snaplist, and can be deleted.
FlashBackup has been changed to remember the snapshot ID between the time
that the snapshot is initially created to the time that the snapshot device
is opened so that it can be deleted.
(NetBackup Clients: Solaris Solaris9 )
--------------------------------------------------------------------------------
Etrack Incident = ET287015
Description:
When running backups with Windows Open File Backup and running multi-
streamed and multiplexed backups, some backups may end with status
41, "network connection timed out". When these errors occur, there should
be multiple (at least 30) log statements such as the one below in the
bpbkar32 logs for the same bpbkar32 process:
12:04:07.788 AM: [4008.3816] <4> V_Snapshot::V_Snapshot_TimeoutCheck: INF -
Attempting to query volume snapshots: bpfis query -id 1103259068
12:04:08.788 AM: [4008.3816] <4> V_Snapshot::V_Snapshot_Destroy: INF -
Attempting to destroy volume snapshots: bpfis delete -id 1103259068
12:04:11.788 AM: [4008.3816] <4> V_Snapshot::V_Snapshot_Destroy: INF -
Snapshot Destroy EXIT STATUS 0: the requested operation was successfully
completed
This is not an issue for non multi-streamed Windows Open File Backup jobs
or backup jobs that have Windows Open File Backup disabled.
Workaround:
To avoid this issue, disable Windows Open File Backup for the client that
have backups end with status 41.
All NetBackup Windows Clients
--------------------------------------------------------------------------------
Etrack Incident = ET287831
Description:
The snap driver has a limit of 32 active snapshots per cache device, which
is not enough in some cases.
Workaround:
To avoid this issue, reduce the file list to use less than 33 file systems,
and create more policies that use different cache partitions.
Additional Notes:
The limit has been increased to 64.
(NetBackup Clients: Solaris Solaris9 )
--------------------------------------------------------------------------------
Etrack Incident = ET284489
Description:
Tar could get into an infinite loop while attempting to restore hard links
when the linked file does not exist or is not a part of the restore job.
Workaround:
To avoid this issue, restore the actual file as well as the hard link, or
make sure the file the hard link is linked to exists for the restore.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET286770 ET216480 ET286770 ET287990
Description:
A VERITAS Volume Manager (VxVM) backup with instant recovery fails when two
file systems are listed. You should not truncate the snapvol name in
pfi_create_snapshot_vol().
(NetBackup Clients: HP Intel-NT Solaris )
--------------------------------------------------------------------------------
Etrack Incident = ET287990 ET216480 ET286770 ET287990
Description:
The following list identifies a few issues with FlashSnap FIM if full-sized
instant snapshots (FMR3 style) are used.
A. You should not use vxsnap split before a disk group split; it deletes
DCO so that FMR will not be utilized.
B. Advanced client does the refresh of the first volume only before a disk
group split; it causes 'fsck' to fail on the alternate client.
C. The calculation of tmp_vol string len is incorrect, resulting in not
enough space being allocated which then causes a core dump.
(NetBackup Clients: HP Intel-NT Solaris )
--------------------------------------------------------------------------------
Etrack Incident = ET288026
Description:
The .f and non-.f files were not getting deleted on a NAS snapshot client.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET289149
Description:
Corrected possible skipped backups if the frequency is a multiple of
24-hours (interval mode) and the backup window is greater than 24-hours.
Only the first backup was run in the failure scenario.
Workaround:
To avoid this issue, make the backup windows no greater than 24-hours.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET289704
Description:
The XBSA Agent PFI backups resulted in status 156 errors. In this
failure, the scheduler is receiving a file list, ?start?(FI_FILE_START)
and file list ?end?(FI_FILE_END), however, it would not receive a file
list ?body?.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET292496 ET289149
Description:
Corrected possible skipped backups if the frequency is a multiple of
24 hours (interval mode) and the backup window is greater than 24 hours.
Only the first backup was run in the failure scenario.
Workaround:
To avoid this issue, make the backup windows no greater than 24-hours.
(All NetBackup Servers)
================================================================================
=========
NB_51_1_M
=========
Etrack Incident = ET185139
Description:
The drives used for disk staging backups are not counted correctly. This
results in under utilization of drives for other backup operations.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET207814
Description:
When scheduler verbose is set greater than 0, there is the potential for
jobs to exit with a status 50. This is more likely to occur on systems that
experience connection issues. The situation occurs when the scheduler
prints a message and loses the errno value. Armed with the incorrect errno
value, the scheduler exits (status 50) the current running job.
Workaround:
Set scheduler verbose to zero.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET206000
Description:
A correction to potential scheduler problems with memory leaks and
non-initialized malloc space has been made.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET205980
Description:
The restart of a frozen image backup job failed. The issue was found using
the following setup:
- VxVM snapshot method to backup a VxFS/VxVM file system.
- checkpoint every 5 minutes.
- scheduled frequency is 1 hour.
- backup attempts are set 2 tries per 1 hour.
The following was the attempted failure scenario:
- Launch the backup and wait for a checkpoint to be taken.
- killed bpbkar processes to put the job into an incomplete state.
- waited for the job is restarted (1 hour later).
- the attempt fails.
- bpfis is attempting to create a new frozen image.
- the previous frozen image still exists.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET207493
Description:
When using the bpexpdate -recalculate or the bpexpdate -backupid commands
to extend the expiration date of one or more backups that have multiple
copies, the media expiration may not be correct for copies on removable
media if the first copy is on disk.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET207498
Description:
You may encounter an occasional failure when duplicating True Image Restore
(TIR) backups. The failure may occur if there is a delay in mounting the
write media.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET167637
Description:
With STK's new version of SSI, certain environment variables are required
to be set when the vm.conf entry ACS_TCP_RPCSERVICE is set. The variables
are set to a default behavior unless they are set differently in the
vm.conf.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET185363
Description:
Corrected compile warnings in the bpadm storage unit free and print
schedule calendar invocations.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET185146
Description:
Bpduplicate, during disk staging duplication, prints the following message
even though no error has occurred.
11/22/2003 11:25:35 AM - Error bpduplicate(pid=1636) from host dcdell1,
COMPLETE_MOUNT QUANTUMDLT70000 0
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET209936
Description:
Resume of a single-stream, suspended job would backup the complete policy
file list instead of only the stream's portion of the file list.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET184689
Description:
In cases where a master was set to require NBAC and a client was
set to automatic without a credential HP-UX would not return the
same error code as other platforms. In such cases error codes 192
and 193 would be documented in the debug log, but would be exposed
back to the user as error code 23.
Workaround:
To avoid this issue, set the bp.conf entry USE_VXSS = PROHIBITED or
USE_VXSS = REQUIRED on client system.
(NetBackup Clients: HP )
--------------------------------------------------------------------------------
Etrack Incident = ET185135
Description:
bpexpdate would not notify the user that they were unable to use bpexpdate
because of access control limiting their access. Even though the binary
returned the appropriate error level (such as, 117, 116, or 118) a human-
readable message was not returned.
Workaround:
To avoid this issue, manually check the error level that bpexpdate returns
(for example, echo %ERRORLEVEL% or echo $?).
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET185137
Description:
Even though the NBU_Operators group did not have backup permission on
Policies, they were still able to perform manual backups. The NBU_Operator
group should be able to do manual backups, however, it should always
require the Backup permission on Policies.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET206796
Description:
Using the options that allow a UNIX file system to be restored without
crossing mount-points fails when restoring all files under the path
instead of just the single file system.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET210609
Description:
A backup with TIR and CPR both enabled in the policy attributes, and files,
instead of directories, specified in the policy file list, can fail with a
status of 130 (signal 11 in the bpbkar log), if the backup goes on for more
than the checkpoint interval.
Workaround:
To avoid this issue, do not enter files in the policy file list, with TIR
and CPR enabled in the policy attributes. TIR works on directories, and so
specifying files in the policy file list is not of much use.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET184717
Description:
Bptm or bpdm may crash after bpsched restarts following a bpsched crash
and subsequently resuming checkpointed backup jobs on media servers. This
change cancels existing duplicate jobs when this situation occurs, or
prevents the crash from occurring.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET210606
Description:
The TLH robotics does not wait for the library auto drive cleaning to finish
before an error occurs and passes off to bptm.
Additional Notes:
In addition, two new syslog entries have been added.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET184971 ET187776
Description:
The VOPIE Enhanced Authentication may not work when host names are greater
than 32 characters. You may see Authentication Failed (error code 160)
messages when connecting with systems with host names or client names
greater than 32 characters.
All NetBackup Windows Clients (All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET184975
Description:
When bpinst -ENCRYPTION pushes the bpkeyutil binary to a UNIX client, the
mode bits of the binary are 0500 on the client. The bits should be 0555.
This could cause a problem if the NetBackup administrator is not running
as "root".
Workaround:
To avoid this issue, login as root on the client and enter this command:
chmod 555 /usr/openv/netbackup/bin/bpkeyutil
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET211098
Description:
The tldcd and tl8cd parent did not check for unit attentions every
15 seconds and caused the AUTO_UPDATE_ROBOT feature to not working properly.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET211433
Description:
When "Browsing for restore", the files that ended with an asterisks '*'
or an "at sign" '@' would have these characters truncated. This prevented
those files from being restored.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET185147
Description:
The following compiler warnings indicate that part of the stack is being
overwritten on these calls.
warning C4133: 'function' : incompatible types -from 'unsigned int *'
to 'time_t *'
Workaround:
This is a rather broad problem that has not appeared in testing, rather as
warnings during a compile. Without these changes, the user should verify
the output of bpcatlist and also verify that bpcatarc is working by
checking the CATARC value after the .f file is archived (of any backup
image that is archived).
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET208248
Description:
Performing backups using an exclude or include list with non-fully-
qualified path names could cause significantly slower backup performance.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = RSVmn50653
Description:
Synthetic Backup fails if the "image copy 1" has been expired, even if a
secondary copy exists.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET213216
Description:
Systems with multiple tld or multiple tl8 robots may see strange behavior
across robot numbers on remote (not control hosts) media servers when the
daemon is bounced. In the worst case, this can include the drives going
into "AVR" control mode on the noncontrol host media servers. Generally,
this is started by a catastrophic failure of one of the robots.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET212553
Description:
Vault currently does not give the user the ability to specify which host
is selected for use as the "robotic control host" for an API robot.
Workaround:
Hand editing the Vault Configuration file
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET209681 ET207797
Description:
Using the command, bpimage -cleanout -client dill-bk, to reduce size of a
database can cause bpimage to core dump.
Workaround:
To avoid this issue, allow the nightly clean-up to clean the catalog.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = RSVmn49450
Description:
NBU_Operators did not have the backup permission on Policies, yet they were
able to perform manual backups. The NBU_Operator group should be able to do
manual backups, but it should always require the Backup permission on
Policies.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET184689 ET229056
Description:
In cases where a master was set to require NBAC and a client was set to
automatic without a credential, HP-UX would not return the same error code
as other platforms. In such cases, error codes 192 and 193 would be
documented in the debug log, but would be exposed back to the user as
error code 23.
Workaround:
To avoid this issue, set the bp.conf entry on the client system to:
USE_VXSS = PROHIBITED or USE_VXSS = REQUIRED
(NetBackup Clients: HP )
--------------------------------------------------------------------------------
Etrack Incident = RSVmn50953
Description:
XBP was inserting and deleting items from lists one item at a time making
the browse for restores very slow. This was done because of deficiencies in
some X-Servers' support of 1.1 Motif. This is not an efficient way to
handle lists with a large number of items.
With X-Servers that support Motif 1.2 this is not an issue. The code was
changed to delete all items at once and to insert multiple items at the
same time.
(All NetBackup UNIX Clients)
--------------------------------------------------------------------------------
Etrack Incident = ET164190 ET164186 ET208178 ET217630 ET218069 ET219861
ET220083 ET223049 ET217198
Description:
Fixed inventory issues for the following libraries:
DELL PV-132T, IBM 3582, Sepaton S100, All STK Libraries, HP A5617A
The added libraries are:
Certance CLL3200
EMC Disk Library
STK SL500
IBM 3590H11
IBM 36232LX
The added drives:
Sony SDX-400V
Added Sony SDX-500V
Certance Ultrium 3
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET213982
Description:
Corrected a restore problem from disk that resulted in a status 92:
- Image contained multiple disk fragments and first file of the
restore was not in the first fragment.
- The size the fragments prior to the one containing the first
file were not a multiple of 1K.
The above conditions can occur during a multiplexed multiple-copy
backup, one or more copies to disk, and one or more copies to tape.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET217243
Description:
Adjusted a loop used during the deletion of a multiplexed group. This
loop was not functioning correctly on the Windows 2000 platform and
resulted in the removal of additional multiplexed groups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET218075
Description:
Bprd hits a segmentation fault and core dumps. The issue appears to occur
when the log file path for the bprestore log exceeds 58 characters. This
only occurs when the media server involved is not the master server, and
when FQDN names are used.
This issue was resolved by increasing the command buffer size to
BUFFLEN (4096), and implemented bounds checking on all string buffers,
where practicable, via the use of V_strlcpy, V_strlcat, and snprintf().
Additional Notes:
While libVcom.so contains no changes, it is required to resolve new entry
points introduced in libsfr.so.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET217923
Description:
Added Solaris10 and Solaris_x86_10 to drop-down menu.
Additional Notes:
For "Solaris Solaris10" to show when configuring clients into a backup
policy, you must execute the following command:
/usr/openv/netbackup/bin/goodies/new_clients
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET206031 ET211885ET211834
Description:
Resolutions to the following issues have been added.
ET206031 Memory leak encountered in the XBSA library.
ET211885 DB2 BLI backup failure
ET211834 BLI backups with multiple streams fail
All NetBackup Windows Clients HP Linux2.4 Linux64 RS6000_433 RS6000_51 SGI
Solaris Solaris9
--------------------------------------------------------------------------------
Etrack Incident = ET211885
Description:
Changes have been added to support DB2 BLI backups.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET219207 ET229830
Description:
Added Solaris10 and Solaris10 x86 to the client drop-down list.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET221820
Description:
There is a potential for missed files if all the following conditions
exist:
1) Advanced client backup.
2) Incremental backup.
3) Multi-stream backup.
4) Backup stream is not stream 1.
Files modified after the snapshot is taken but before the (non-stream 1)
image is created may be missed.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET219245
Description:
After an NDMP import, the file count drops by a very large number.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET225034
Description:
An incremental multi-stream snap backup with forced stream 1 can miss files.
The problem exists when the following conditions are true:
- Snap backup (advanced client, WOFB).
- Incremental backup.
- Multi-stream backup.
- Stream 1 is not due, but is force-started to perform the snapshot
operation. Aggravated by ?frequency? mode (frequency is NOT a multiple of
24-hours).
The problem can be minimized/eliminated by configuring for an "all stream"
backup:
- Use calendar-based scheduling.
- Use ?interval? mode (frequency is multiple of 24-hours).
- Avoid multi-streaming snap backups.
This problem occurs because of an oversight in cloning a due stream into
a non-due stream1. Remember, stream1 is not due, but is required to
perform the snap shot operation. The due stream information is copied in
mass to create the stream1 information, and then adjusted to make it
stream1 unique. The error occurs because the backup times (last backup,
last full backup, last incremental backup) were not updated to the correct
stream1 values. As a result of this, stream1 uses the stream2 image
creation times. Image creation times are used to compute the delta time
which determines the targeted incremental backup files.
Unfortunately stream2 also suffers the same failure. The stream2
incremental backup should be based upon the stream1 snap time, but stream1
is using the incorrect time for the snap. So, all the due streams suffer
this failure.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET219966
Description:
Resume of a checkpoint backup could select a different media server if a
failure on initial backup attempt occurred.
The failure scenario is:
1) Backup reached a checkpoint.
2) Then failed (this case was a 13 "read failed").
3) Scheduler attempts to programmatically re-submit.
4) But the drives were re-set & downed.
5) So, the job exited with a 213 "no storage units available".
The original storage unit information was "lost" so the job could
potentially resume on a different media server. If resumed in this way,
the backed up data would reside on two different media servers and any
subsequent restore would fail.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET229834
Description:
A loss of multiplexed groups within bpsched.exe resulted in the failure to
cancel multiplexed backups.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET229949
Description:
Unable to restore using individual file restores through FlashBackup on a
Windows client with an AIX master server, if a zero length file exists in
the restore list.
Workaround:
Restore will work if no zero length file is included in the restore list.
(All NetBackup Servers: RS6000)
--------------------------------------------------------------------------------
Etrack Incident = ET228739
Description:
Listing the master in the ClientList no longer causes update_clients to
fail. To correct this issue, a check was added in bpinst so that if the
client being shipped to is the master, and to not try and ship the java tar
ball, which is very large. It will also corrupt the file, and it is
already in the right location. In addition, a change was made to
update_clients in the case where the master is the client as nbj.conf has
moved to nbj.conf.bak.
(All UNIX NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET229844
Description:
An intermittent problem with installer no longer occurs when trying to
install to remote clients. The following two files were not being copied
to %TEMP%\X86, and caused the following error to appear, "The system
cannot find the file specified" on Install GUI.
VERITAS Maintenance Pack.msi
Data1.cab
Workaround:
To avoid this issue, do the following:
1) Start the installer. Select "Select from available computer on the
network". Click Next.
2) Copy VERITAS Maintenance Pack.msi Data1.cab to the %TEMP%\X86 directory.
(easiest way to find what %TEMP% is, is to open a cmd shell and run echo
%TEMP%, usually its something like C:\Documents and
Settings\<UserName>\Local Settings\Temp\X86).
3) Proceed with install.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET230865
Description:
An initialization failure against nbwin.exe is generated when a
non-administration account initiates a restore. A pop-up is generated
stating that nbwin.exe failed to initialize 0xc0000022. A change to the
permissions was made to msvcr71.dll so that the user "Users" has Read and
Execute permissions.
Workaround:
Manually change the security permissions on msvcr71.dll to allow Read and
Execute permissions for "Users" user.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET211802
Description:
The cancel_all_exceeded_max_wait() function does not scale well in large
environments. This change modifies the way incomplete jobs are processed
when they have exceeded the maximum incomplete time.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET224678
Description:
Improved VMD debug logging to better enable analysis for slow servers.
Allow control of VMD child timeouts, such that slow servers can be
interrupted, before VMD encounters connection timeout's on other
connections.
Add "VMD_CHILD_TIMEOUT = xx", to the vm.conf file. The value "xx" is the
number of seconds to wait for a child to release its lock on the volume
database. Allowable time is 60 - 360 seconds, where 360 is the default
(and value used prior to this patch).
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET218653
Description:
A race condition existed when two restore jobs needed the same piece of
media, from the same server, when this media was not already in the media
database. The condition could cause the media to be added twice and one
of the restores to fail.
(All NetBackup Servers)
--------------------------------------------------------------------------------
Etrack Incident = ET231394
Description:
Oracle RMAN backups taken using UNIX Media servers fail with a status 25.
(All UNIX NetBackup Servers)
===============================================================================
Attachments
|
|
Related Articles
Legacy ID
280422
Article URL http://www.symantec.com/docs/TECH45020
Terms of use for this information are found in Legal Notices
Thank you.