On Tue, Oct 14, 2008 at 02:36:40PM -0700, Adams, Dwayne wrote:
> I have 2 NDMP filers that die with error 23s at approximately 15 minutes
> into FULL backup  jobs.   I believe Netbackup is killing the NDMP
> connection after 900 seconds (waiting for a data response from the
> filers).

Pretty odd.  I wouldn't expect timeouts on a FULL.  Anything in filer
backup logs?  

Do you have a netbackup/db/config/NDMP_PROGRESS_TIMEOUT file?

> Are some timeout settings combined during the backup process?
> For instance, will a combination of client connect timeout (300) and
> client read timeout (600) equal a 900 timeout and cause Netbackup to
> kill the session at 900 seconds? 

I'm not even certain those settings apply to NDMP connections.

> 10/11/2008 6:23:28 AM - Error ndmpagent(pid=3848) Connection was closed
> but has not yet been destroyed.    

That's odd...  I'd really be interested in the filer backup log side.
Any chance there's a firewall/nat shutting down the control connection
between the NBU and filer?

-- 
Darren
_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu