backup exec services stop during backup job

[charles_collett]

We are having an issue when running a duplicate backup job to tape. the original backup goes to disc and works just fine - then the duplicate job to tape starts without any problems, but during this job all the BE services die killing the job. In Job History I can see the 'Job History' tab, but when I click on the 'Job Log' tab the services die again and it displays nothing. There are no alerts so I'm confused as to why this is happening? It seems to occur towards the end of the job and from what I can tell the job actually completes - judging by the amount of data on the tape afterwards. We have other duplicate jobs to tape that are completing just fine, this is the only one giving us trouble. This backup job is backing up the passive node in an Exchange 2007 CCR Cluster. It has worked fine in the past but just started acting up this week weekend. We are using no encryption and are using BE 11d with all the latest updates.

Any ideas?

[Ben_L_]

First thing you want to look at is your event logs. check the application and system logs for errors around the time the service stopped.  This may help us to figure out what is going on.

[charles_collett]

Thanks for the help. I have recreated the issue and then checked the event viewer as you suggested. here is that I'm getting.

Application:
Faulting application beserver.exe, version 11.0.7170.27, faulting module beserver.exe, version 11.0.7170.27, fault address 0x00000000001f3458.

System:
The Backup Exec Server service terminated unexpectedly.  It has done this 2 time(s).

Not sure what else to check or if this gives us any clues as to why the services keep stopping during this backup job. In fact if I simply try to look at the Job Log on this specific backup job it kills as the services...

thanks.

[Ben_L_]

Interesting, was hoping for a better error.  But we can work with this.  Normally when I see beserver crash it has something to do with .NET 2.0.  Now I know MS just released some updates for this, can you try running a repair of .NET or uninstalling and reinstalling .NET 2.0?

[charles_collett]

Thanks for the reply. As you suggested we first ran a repair operation on .NET 2.0. After the repair the problem in BE persisted so we went ahead and removed .NET 2.0 and then reinstalled. Again the problem continues. Furthermore when I log on to this server I am now seeing the following error pop up:


Backup Exec RPC Server
szAppName : beserver.exe     szAppVer : 11.0.7170.27    
szModName : beserver.exe     szModVer : 11.0.7170.27    
offset : 00000000001f3458    


Do you have any other suggestions?

[Ben_L_]

Have you tried running a repair of Backup Exec or using BEUtility to run a repair of the database?

[charles_collett]

Thanks for the suggestions again. We went to add/remove programs and ran a repair operation on BE. Then also went into BEUtility and repaired the database. However the problem still persists..... I'm really not sure what to do at this point.  Could it be specific to this one backup job, since my other duplicate to tape jobs run fine?

[Rustylinux]

Hello,

Some questions I would ask are:

1.)  Does this happen with any tape?

2.)  Does this happen with any backup ? and does the size of the backup matter?

3.)  Is your backup server taxed for memory during a backup process ??

4.)  Is your tape drive on a scsi controller that is properly terminated and has no other devices sharing the controller.  Sometimes when you have 2 devices sharing the same controller time out can occur.

5.)  Is your backexec patched right up to date ?

6.)  You could also try .Net 3.0 if your out of ideas.

7.)  I also found this from another post with the same problem:

"I would suggest upgrading to the latest revision of 11d.  You are currently running the first rev (6235).  If you want to get the latest rev, just go to fileconnect.symantec.com or backupexec.com and download the latest installer.  It will upgrade your current install and you should not have to recreate anything."

I hope some of these things help you find a resolution.

[charles_collett]

Thanks for your suggestions.

1) no - this only happens on one specific job, other jobs to tape are fine
2) backup size is not an issue as this one is smaller than other successful jobs to tape
3) don't think it's with the hardware. This is a dedicated backup server with two quad core xeons and 4GB of ram - BE is the only app running.
4) again - other jobs to tape run fine
5)BE is current:
Media Server: Version 11.0 Rev. 7170
and we have run liveupdate with all current patches as fo yesterday including service pack 2
6) I'll install .Net 3.0 and see, but again this is specific to this job for some reason - other duplicate to tape jobs run fine so I don't see how it is a .NET issue

7.) Unfortunately this doesn't apply as our BE 11d is up-to-date

What is sketchy is that when I go to simply look at the Job History from this failed job, just by doing this I get the message saying "unable to retrieve job log. No further information available" and it kills all the BE services. Now I can see information in the "Job History" tab, but nothing in the "Job Log" tab.

[Rustylinux]

Hey,

It sounds to me as if that job is just corrupted for one reason or another.  Was this job transfered from an older job or version of Backexec?? if so I would just manually re-create the whole job.

Also if you where to run the job without it being part of a duplicate does the same error still occur ??

Also this may be a stupid question but does your backupexec user account that you use to do the backups have permissions to ALL the files you are trying to backup with this one job.  I have seen it where for some stupid reason the Backupexec user account the user is using to backup doesn't have permissions on a few files or locations in the backup and the job will failed and bail out of the job instantly when this happens with little to no proper error code pointing to this.  Make sure your Backup exec user account that you use to backup is a domian admin and that all the files you are trying to get in the duplicate have sufficent permissions for Domain admins or that user however you have it setup.

Ohhh also is the tape drive you have on its own scsi card ? or is it shared with your hard drives or other scsi devices ???  If you had a scsi card controlling your hard drives and your tape drive on a different channel this sometimes causes jobs to bail out with stupid error codes.

Message Edited by Rustylinux on 12-28-2007 12:44 PM

[charles_collett]

Thanks for the tips again. The only thing left I can see is that yes this specific job must be corrupted some how. The autoloader has its own scsi card and I've been using a domian administrator account with sufficient priveledes for all the jobs but this is only one with a problem. And again the job runs fine, but then when the 'duplicate to tape' job launches is when the problems start. I'm going to blow away this backup job and start fresh and see what happens.