We are finding that our restores are painfully slow after upgrading to 11d. Our backups are running at the same speed they were before the upgrade.
It appears that the tape unit is just seeking forever and ever. If we choose to redirect the files to another location instead of restoring them to the original, things move faster.
To qualify "slow" and "fast", I've run the same job twice, restoring a 34MB Access database; once with redirection off and it didn't write anything after 3.5 hours (I had to cancel it before it tied up our tape unit for so long that it started impacting our nightly backups); and again with redirection on and it finished the job successfully in 37 minutes. I've tried re-creating the job, restoring files on different file servers, and the behavour is consistent.
I am trying to see if Preserve Tree has anything to do with it, but so far it doesn't appear that it does.
Has anyone else seen this one? What could be causing the tape unit to seek/read endlessly for no apparent reason?
I've read
http://support.veritas.com/docs/287309 and got a chuckle out of the "
Note: This issue is not data threatening. The data can be restored it just takes longer than expected." and "Let the restore read through all of the media and complete normally. This will add time to the restore but it will finish."
Really? Why is a 3+ hour restore for a 34MB file acceptable with a solution of "It's fine, just wait it out."
