File backup-Recurring error 52

[neetu]

Hi all,

I had this stuation several times where the incremental file backup fails with error 52 however many times I try to restart or resume it. However, the same client backs up successfully if full backup is triggered. Does anyone have an explanation for this?

[Marianne]

Is it the same media id used each time for this incremental backup?

What is media mount timeout set to?

Have you checked bptm log on the media server for the reason why tape cannot be mounted?

[neetu]

No. The error recurred on different media. Also, same media was used for the full backup later. No issues found then!

As of now I do not have the bptm logs handy.

[sri_vani]

Please provide the logs as Marianne suggested and verify the Media Id

1)what kind of backup it is?

Request you to verify similar kind of issue?

http://www.symantec.com/business/support/index?page=content&id=TECH50982

[Marianne]

Please create bptm log folder on the media server.

There is no way to troubleshoot this without logs.

Please also copy all text in Details tab of failed job and post here.

You also forgot to tell us what Media Mount Timeout is set to.
I always recommend this timeout to be no less than 20 minutes.

[neetu]

Hi Marianne,

Media mount time out is set to 15 mins in the environment.
 

Just to make things clear, the incremental backup does not fail everyday. Once in a while a situation occurs where incremental backups keep failing and at such times if full backup is triggered, it completes successfully. The incremental backups following this full backup, complete successfully without any issue.

Detailed status of the failure attached.

[Marianne]

You never mentioned your NBU version?

I remember a bug some time ago where a different kind of timeout (e.g. Client Read timeout) is incorrectly reported as status 52.

I see the following in Job details:

23/10/2013 14:17:48 - Info bpbrm(pid=9020) Sending the file list to the client     
23/10/2013 14:32:49 - Info bpbrm(pid=8800) child done, status 52   

15 minutes between the 2 timestamps - the value of Media Mount timeout when it actually looks like Client Read Timout (no data received from client since last timestamp). Client is probably taking a long time 'walking' the filesystem to find changed files.

This is after tape was mounted, positioned and waiting for data:

23/10/2013 14:16:14 - Info bptm(pid=9124) Waiting for mount of media id X10667 (copy 1) on server mno.
23/10/2013 14:16:15 - Info bptm(pid=9124) media id X10667 mounted on drive index 29, drivepath {4,1,14,0}, drivename <DRIVE NAME>, copy 1
23/10/2013 14:17:02 - mounting X10667
23/10/2013 14:17:02 - mounted; mount time: 00:00:00
23/10/2013 14:17:02 - positioning X10667 to file 596
...

23/10/2013 14:17:47 - positioned X10667; position time: 00:00:45
23/10/2013 14:17:47 - begin writing

So, going forward, I suggest you do the following:

Increase Media Mount timeout to 45 minutes.
Increase Client Read Timeout to 30 minutes (1800).

Create all of these log folders for troubleshooting should the issue persist:

On media server: bpbrm and bptm

On client: bpbkar

Hope this helps.

[watsons]

http://www.symantec.com/docs/HOWTO34938

There is a possible cause that worths checking in this technote:

the last available drive has a mount request for a non-backup (such as a restore). Then a backup that requires the same drive is initiated before the mount completes. This error is due to the drive not being reported as busy until the mount completes.