Veritas Cluster Server Management Console 5.1 Patch 1144918d for Solaris

Article:TECH59649  |  Created: 2008-01-15  |  Updated: 2009-01-15  |  Article URL http://www.symantec.com/docs/TECH59649
Article Type
Technical Solution


Environment

Issue



Veritas Cluster Server Management Console 5.1 Patch 1144918d for Solaris

Solution



Patch for Veritas Cluster Server (VCS) Management Console 5.1
========================================================================
Name:                           5.1+e1144918

Date:                           07 Feb 2008

VCS Management Console Release:    5.1


PROBLEM DESCRIPTION:
--------------------

This point patch addresses the following known issues in the 5.1
release of the VCS Management Console:

* Management server crashes after creating report which has Japanese
 text in "Report Description." (1132801)

* Management server runs out of memory for large environments. (1142963)

* If a cluster is deleted from the local management server, the cluster
 is still present in the peer management server. (1131263)

* Online Help does not display .png files. (1134337)

* The date field is not internationalized under the Log tab. (1141872)

* The "Assign Privileges on Clusters" wizard does not remove old
 privileges when assigning new ones. (1121315)

* Adding remote clusters fails for certain secure clusters. (1129642)

* It takes a long time to log on to the console for large environments.
 (1144143)

* "Enterprise -> All Users -> All VCS Users" page takes a long time for
 large environments. (1144143)

* Some cluster logs may be lost on upgrade. (1134936)

* Error in reconnection logic when cluster restarts while Cluster
 Connector is waiting to synchronize with the management server.
 (1143714)

* Some user privilege assignment operations may fail for clusters
 managed by a peer management server. (1129489)

* Console will show Japanese text in some cases even if the user logs
 in with the English locale. (1133894)

Apply this point patch on the management server. If you want to address
the issue of "Adding remote clusters fails for certain secure clusters.
(1129642)" then apply part of this point patch on the VCS cluster that
is experiencing the problem.

For the latest information on updates, patches, and software issues
regarding Veritas Cluster Server Management Console, see the following
TechNote on the Symantec Technical Support web site:


 http://entsupport.symantec.com/docs/290657

 
FILE AFFECTED BY THIS PATCH
---------------------------

* Management Server:

 Solaris:
     /opt/VRTScmcm/webapp/cmc.war

 Windows:
     C:\Program Files\Symantec\VRTScmcm\VRTSweb\admin\root5\ROOT5.war
   
* Cluster Connector (For 1129642 only):

 All Unix platforms:
     /opt/VRTScmccc/lib/Common.jar

 Windows:
     C:\Program Files\Symantec\VRTScmccc\lib\Common.jar


VERIFYING CRC AND BYTE COUNT
----------------------------

Use the cksum command to ensure that the file you have downloaded
matches the following cksum and byte count:

# cksum cmc.war Common.jar

364612631      28110123        cmc.war
632094981      940655          Common.jar

The output displays the CRC value (first value), the byte count
(second value), and the file name (third value).


PATCH INSTALLATION INSTRUCTIONS
-------------------------------

DOWNLOAD THE PATCH:

1. Click the Download link below.

2. On Solaris, extract the files from the archive using the following command:

tar xvf 1144918d_302919.tar

The following files are extracted:
cmc.war
Common.jar
README.TXT
webgui.exe

3. Then follow the installation instructions below.

Solaris:
--------
1. Run the following command to verify that the system has the correct version
  of the management server installed:

  # pkginfo -l VRTScmcm

  For example:

  # pkginfo -l VRTScmcm

  PKGINST:  VRTScmcm
     NAME:  Veritas Cluster Server Management Console by Symantec
 CATEGORY:  application
     ARCH:  sparc
  VERSION:  5.1
  BASEDIR:  /
   VENDOR:  Symantec Corporation
     DESC:  Veritas Cluster Server Management Console for multiclust...
   PSTAMP:  paladin20070917233602

2. Stop the management server.

  For a standalone management server, use the following command:

  # /opt/VRTScmcm/bin/vxcmcweb stop

  For a clustered management server, use the following command:

  # hagrp -offline CMC_MS -any

  If this command times out, you can safely kill the relevant processes using
  these commands:

  # ps -ef | grep CMC | grep -v grep
  # kill <java_pid> <db_pid>

  For example:

  # ps -ef | grep CMC | grep -v grep

  root 15905     1  0   Nov 02 ?       137:15 dbsrv9 -n CMC_VCSLABWEB_14994 -x.
  root 15844     1  0   Nov 02 ?       114:14 /opt/VRTSjre/jre1.5/bin/java -D..

  # kill 15905 15844

NOTE: For the clustered management server environment, repeat steps 3-5
     on each clustered system.

3. Use the following command to make a backup of the original cmc.war file:

  # cp -p /opt/VRTScmcm/webapp/cmc.war /opt/VRTScmcm/webapp/cmc.war.orig

4. Use the following command to delete the previous ROOT5 folder:

  # rm -rf /opt/VRTScmcm/VRTSweb/admin/root5/ROOT5

5. Use the following command to replace the old ROOT5.war file with the one
  provided in this point patch:

  # cp -p <patch_folder>/cmc.war /opt/VRTScmcm/webapp/cmc.war

6. Use the following command to change the garbage collection settings for java:

  # /opt/VRTScmcm/VRTSweb/bin/webgui opts "-XX:NewRatio=5"

7. If you plan to manage more than 25 clusters using the management server, use
  the following command to give more memory to the java process:

  # /opt/VRTScmcm/VRTSweb/bin/webgui maxheap 1500

  NOTE: This assumes you have at least 2 GB ram on your server. The argument
  "1500" to the command above is in MBs. You can give more or less ram to your
  java process based on your environment. Setting value to 1500 will let you
  manage up to 15,000 cluster objects, which is equivalent to about 75 clusters,
  each with 2 systems, 10 service groups, and 100 resources.

8. Restart the management server.
 
  For a standalone management server, use the following command:

  # /opt/VRTScmcm/bin/vxcmcweb start
 
  For a Clustered management server, use the following command:

  # hagrp -online CMC_MS -any

9. Wait for several minutes and then use the following command to confirm that
  the server has restarted:

  # ps -ef | grep CMC | grep -v grep

  The output should show two running processes.


For 1129642 only
----------------

On each cluster with systems requiring the fix for #1129642, do the following:

1. Use the following command to take the CMC group offline and stop the Cluster
  Connector:

 #hagrp -offline CMC -any

2. On each cluster system, use the following commands to backup the existing
  Common.jar file:

  # cd /opt/VRTScmccc/lib
  # mv ./Common.jar ./Common.jar.orig

3. Use the following command to replace the Common.jar file with the Common.jar
  included in the hotfix archive:

  # cp -p <patch_folder>/Common.jar /opt/VRTScmccc/lib/Common.jar

4. On any cluster system, use the following command to bring the CMC group
  online and start the Cluster Connector:

  # hagrp -online CMC -any

5. Verify the cluster has reattached to the Management Server by checking the
  status on the Management Server web console.


DISCLAIMER
This is a supported fix that has not been fully regression tested and must
be applied only to systems that are severely affected by the problem. For
other situations, we recommend that you wait for the next Maintenance
Pack that will contain this fix. Contact Symantec Technical Support for
more information.


Attachments

1144918d_302919.tar (28.5 MBytes)


Legacy ID



302919


Article URL http://www.symantec.com/docs/TECH59649


Terms of use for this information are found in Legal Notices