|
Re: [Veritas-bu] Sudden DSSU Flush problems
2010-08-06 10:02:55
On Thu, Aug 5, 2010 at 9:26 PM, Nate Sanders <sandersn AT dmotorworks DOT com> wrote:
- 4 luns coming off a NetApp 6080 to the RHEL Master NBU 5.1MP6 server
(dssu1,2,3,4)
- all DSSU's 700GB
- dssu3/4 in a group for Oracle RMAN Archive logs
All of a sudden this week we're seeing the disk staging flush (relocate
to final destination) jobs taking 2x-3x as long as they should. Lately
we've been having a lot of 84 (media write errors, which in 5.1 on DSSU
means it's full) errors on DSSU 1 which stores OS backups. I kept trying
to work with the schedule but it seemed whether it was 4hrs or 6hrs it
would always error out about %50 of the clients
Status 84 could also mean you have hardware problems - I've certainly seen this before with flaky fibre paths.
Have you checked the OS server logs or done non-NetBackup read/write tests to prove the hardware is really operating as it should be? What happens if you try and tar up the files and send them to the null device - do they go as fast as they should or are they slow too?
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
|
|
|
