NetBackup upgrade failed

[liuyang]

Hi, we are helping a customer to upgrade their NBU servers. Their master server and media servers are all NBU 7.5.0.3 on Windows Server 2008 R2 SP1. We helped them to upgrade to NBU 7.5.0.7. The upgrades of media servers have no issues.

During the upgrade of master server, we saw an error message which said "error in updating NetBackup database". But the upgrade still completed. However, after the upgrade, we found the NBU relational database service failed to start. I tried to run "nbdb_admin -validate -full" command to check and it shows Database [NBDB] is not available. In the upgrade log, it has an error which said "assertion failed: 201001, collation is corrupt". 

Does anyone have any idea regarding this? Thanks in advance.

[Marianne]

Log SEV 1 call with Support.

We can ask for logs, but it will take time and most of us have limited time... 

[Will_Restore]

SImilar message, http://www.symantec.com/docs/TECH205648

-edit-

best to log a case per Marianne to be safe

[Marianne]

Hope you have logged a call by now and that you are busy talking to a Support engineer....

If you have a moment, please post nbdb and server logs here:

<install_path>\VERITAS\NetBackup\logs\nbdb
<install_path>\VERITAS\NetBackupDB\log\server.log

Please copy logs to .txt files and upload as File attachments.

Emm log may also help:

vxlogview -o 111 -t 01:00:00
(above command will collect logs for the last hour. Adjust the time to include time NBU attempted to come up after the upgrade)
or:

vxlogview -o 111 -b <StartDate> -e <EndDate>
(start and end dates must be in this format: "1/1/2013 12:00:00 AM")

[liuyang]

Thanks for your replies.

We logged a case with Symantec support already. We are able to start the NBU services after recreated NBDB database. However, after that all the settings are gone. We are now in the process of recovering database.

[Marianne]

Hopefully up and running by now?

Have you been able to find the reason for the failed upgrade?

[liuyang]

Thanks very much for all your replies. 

With the help of Symantec support, we managed to start the master server by re-created the NBDB databases and then recovered catalog.