RMAN backups failing.

[Rocky67]

Netbackup logs and acticity monitor logs show backup as successful, though data written is very less..few kbs.. alert generated as backup failed :

RMAN-03009: failure of backup command on t2 channel at 01/02/2016 06:10:48

ORA-27192: skgfcls: sbtclose2 returned error - failed to close file

ORA-19511: Error received from media manager layer, error text:

   VxBSAEndTxn: Failed with error:

[Marianne]

Please post full rman output file as well as dbclient log. Copy logs to .txt files and upload as File attachments.

[Rocky67]

uploaded RMAN error log

[Marianne]

Please post dbclient log as well. If the log folder does not exist, create it under /usr/openv/netbackup/logs with 777 permissions. Post this log after next failure. Please post text in Activity Monitor child job(s) as well. It seems weird that NBU is reporting success but rman shows a failure.

[Rocky67]

Uploaded dbclient logs

[Rocky67]

activity monotor job details posted

[Marianne · Accepted solution]

This looks like output for a single child job. Surely there are more child jobs for Oracle backup? Client_Read_Timeout on media server for large dbs should be more than 300. The recommended timeout value for Oracle backups is 1800. Please ensure that dbclient log folder exists on the client as well as bptm and bpbrm on the media server.

[Michael_G_Ander]

Agree with Marianne, the error is probably caused by a timeout.

For an incremental backup it can be caused by the Oracle database scanning for changed blocks, one possibile migation of this is to use Block Change Tracking.

For NetBackup to "know" that rman failed to exitcode needs to be send back to NetBackup, think it is still explained in the Oracle guide for Unix and Windows

[Marianne]

 No response in recent days....

I found some info on the web that may help:

https://www.veritas.com/community/forums/oracle-backups-failure-and-archive-log-backups-failed-error... 

http://www.veritas.com/docs/000020638 

http://www.dba-oracle.com/t_ora_27192.htm 
(This one is for different backup product but the solution was with network settings on the Oracle client.)

[Rocky67]

Thank you Marianne.

Issue is resolved. Storage server was processing jobs slow. As a result job was queued for long time causing RMAN timeout ,  the long queued job when started writing had a timeout., . Storage server rebooted and its processing fine....