System State RC 13 wk8

[FriZY]

Hello,

we have a problem with backup of system state on windows 2008 64bit . Backups always crashed with return code RC 13.

I tried to disable and enable WOFB and change VSP to VSS but nothing help.

Could you help me with this issue? THX

master: Solaris 10 Netbackup 6.5.5

client: Wk8 64 bit Netbackup  6.5.5

Log:

8.3.2011 21:09:08 - Error bptm (pid=26625) system call failed - Connection reset by peer (at child.c.1296)
8.3.2011 21:09:08 - Error bpbrm (pid=26626) socket read failed: errno = 131 - Connection reset by peer
8.3.2011 21:09:08 - Error bptm (pid=26625) unable to perform read from client socket, connection may have been broken
8.3.2011 21:09:09 - Error bpbrm (pid=26626) could not send server status message
8.3.2011 21:09:11 - end writing; write time: 0:04:36
file read failed (13)

[Marianne]

Please create bpbkar and bpfis log directories on this client.

These logs will give more information and will assist with troubleshooting.

This TN describes a bug that was supposed to be fixed with NBU 6.5.3: http://www.symantec.com/docs/TECH62588

Another TN: http://www.symantec.com/docs/TECH87247

[FriZY]

@Marianne

THX

When we switch cluster then backup finished successfully but on inactive node backup crashed with same RC.

So problem is how to backup SYSTEM STATE on inactive node of cluster.

[SteveYu]

My personal opinion on this is that something with the VSS components on that one node are broken.

1. I would do check VSS components by running "vssadmin list writers" from the broken node and the working node and compare them.

2. You should also check for errors in the windows event logs

3. Have you ever been able to backup the shadow copy components of the broken node?

4. On the broken node check in the Windows Services for "Microsoft Software Shadow Copy Provider" and "Volume Shadow Copy". On a working system they should be Blank status and manual startup type. Additionally you can compare this against the working node. 

I have seen that in some cases the VSS services were in a hung state and simply putting them back into the right state corrected the problem

5. The super test is to remove Netbackup from the problem. You can try to use Windows server backup (Ntbackup) to backup the system state to see if that works. If Ntbackup can't back up the system state then of course Netbackup can't

-sy

[FriZY]

thx for hint,

Yes, backup of system state was working on both nodes till now. We have currently problem only on inactive node of cluster. Even if we switched cluster problem still occured on inactive node.

[FriZY]

We found out TECHNOTE for such problem.

http://www.symantec.com/business/support/index?page=content&id=TECH135755

So, I hope that this patch solve my issue.

[Rajesh_s1]

You need to install the hot fix and exclude the paths for the registry . Just installing hotfix will not fix this issue .

We also had the similar problem with the passive cluster nodes where SAP is installed . It was due to some of the SAP components were included in system state component . So that SAP services which was active in active node and was not accessable at passive nodes so it was giving error at passive node .

We have foundout that paths and excluded from the system state backup and installed the hotfix after that it started working fine ...

Please specify your application running on win 2k8 machine.