Problems restoring a Solaris 10 Update 6 (10/08) or later Bare Metal Restore client
| Article:TECH69398 | | | Created: 2009-01-18 | | | Updated: 2009-01-18 | | | Article URL http://www.symantec.com/docs/TECH69398 |
Problem
Problems restoring a Solaris 10 Update 6 (10/08) or later Bare Metal Restore client
Error
Boot load failed.
The file just loaded does not appear to be executable.
Solution
Due to extensive changes made by Sun with this Solaris
release, the Bare Metal Restore (BMR) environment must be upgraded in order to
properly restore a client backup image created at this OS release
level.
A client at this OS release level can only be recovered using a Shared Resource Tree (SRT) at the same level or higher. However, the released software for BMR does not currently allow for the creation of a SRT at this level. New binaries are available and must be installed in order to perform this function. Please specify ETrack 1473582 for release 6.5.2, ETrack 1538448 for release 6.5.3 or Etrack 1703721 for release 6.5.4.
To completely incorporate this fix, the BMR Boot Server itself must be upgraded to Solaris 10 Update 6 or later level as well.
A Solaris 10 Update 6 or later SRT can only be used to restore Solaris 10 Update 6 or later client images. Earlier versions are incompatible.
The same is true for SRT versions at Update 5 and earlier, which can not restore an Update 6 or later client image.
If a Solaris 10 Update 6 or 7 client is restored using an SRT based on an earlier Solaris install media Update 5 or earlier, the process will typically run to completion without error. However, because the boot mechanism was changed starting with Update 6, the client will not properly reboot.
A typical boot failure message on the console will be:
Boot load failed.
The file just loaded does not appear to be executable.
A possible work around exists to allow for the client to reboot.
1. Boot the client, in single user mode, from the same media or network image that was used to recover it.
2. From the command line create a mount directory off the /tmp directory
3. Mount the slice(s) that constitute the root (/) file system as well as the /usr file system, if separate.
4. Run the install bootblock command from the overmounted file systems, specifying the correct target disk.
5. Sync the files systems and reboot.
As an example:
1.)Starting at ok prompt do network boot into single user mode:
ok> boot net -S
2.)Mount file systems:
# mkdir /tmp/mnt
# mount /dev/dsk/c0t0d0s0 /tmp/mnt
# mount /dev/dsk/c0t0d0s5 /tmp/mnt/usr
3.)Re-create the boot block:
# /tmp/mnt/usr/sbin/installboot /tmp/mnt/usr/platform/`uname -i`/lib/fs/ufs/bootblk /dev/rdsk/c0t0d0s0
(NOTE: the two (`) characters in /`uname -i`/ portion of path above is a back tic, which is the key left of the "1" key on your keyboard ).
# cd /
# umount /tmp/mnt/usr; umount /tmp/mnt
# sync; sync; sync; reboot
You can determine the boot image kernel level by looking at console output. As an example:
SunOS Release 5.10 Version Generic_127127-11 64-bit
The kernel level noted is 127127-11, which is Solaris 10, Update 5, with patches.
For Solaris 10 Update 6, the kernel level will be Generic_137137-09 or higher.
A client at this OS release level can only be recovered using a Shared Resource Tree (SRT) at the same level or higher. However, the released software for BMR does not currently allow for the creation of a SRT at this level. New binaries are available and must be installed in order to perform this function. Please specify ETrack 1473582 for release 6.5.2, ETrack 1538448 for release 6.5.3 or Etrack 1703721 for release 6.5.4.
To completely incorporate this fix, the BMR Boot Server itself must be upgraded to Solaris 10 Update 6 or later level as well.
A Solaris 10 Update 6 or later SRT can only be used to restore Solaris 10 Update 6 or later client images. Earlier versions are incompatible.
The same is true for SRT versions at Update 5 and earlier, which can not restore an Update 6 or later client image.
If a Solaris 10 Update 6 or 7 client is restored using an SRT based on an earlier Solaris install media Update 5 or earlier, the process will typically run to completion without error. However, because the boot mechanism was changed starting with Update 6, the client will not properly reboot.
A typical boot failure message on the console will be:
Boot load failed.
The file just loaded does not appear to be executable.
A possible work around exists to allow for the client to reboot.
1. Boot the client, in single user mode, from the same media or network image that was used to recover it.
2. From the command line create a mount directory off the /tmp directory
3. Mount the slice(s) that constitute the root (/) file system as well as the /usr file system, if separate.
4. Run the install bootblock command from the overmounted file systems, specifying the correct target disk.
5. Sync the files systems and reboot.
As an example:
1.)Starting at ok prompt do network boot into single user mode:
ok> boot net -S
2.)Mount file systems:
# mkdir /tmp/mnt
# mount /dev/dsk/c0t0d0s0 /tmp/mnt
# mount /dev/dsk/c0t0d0s5 /tmp/mnt/usr
3.)Re-create the boot block:
# /tmp/mnt/usr/sbin/installboot /tmp/mnt/usr/platform/`uname -i`/lib/fs/ufs/bootblk /dev/rdsk/c0t0d0s0
(NOTE: the two (`) characters in /`uname -i`/ portion of path above is a back tic, which is the key left of the "1" key on your keyboard ).
# cd /
# umount /tmp/mnt/usr; umount /tmp/mnt
# sync; sync; sync; reboot
You can determine the boot image kernel level by looking at console output. As an example:
SunOS Release 5.10 Version Generic_127127-11 64-bit
The kernel level noted is 127127-11, which is Solaris 10, Update 5, with patches.
For Solaris 10 Update 6, the kernel level will be Generic_137137-09 or higher.
|
|
Related Articles
Legacy ID
322801
Article URL http://www.symantec.com/docs/TECH69398
Terms of use for this information are found in Legal Notices









Thank you.