Legacy HDS Forums

Performance impact of zero-scrub operation on USP-VM external Tier-2 disk

Discussion created by Legacy HDS Forums on Mar 11, 2012
Latest reply on Mar 12, 2012 by Legacy HDS Forums

Originally posted by: minderbin


When dis-associating DP V-VOLs from DP-Pools and subsequently deleting the V-VOL for any V-VOLs utilising external Tier-2 disk I'm experiencing a substancial performance hit for production hosts utilising storage in the same DP Pool (very high disk response times).

In monitoring the external array, I'm seeing very high write utilisation on the volumes comprising the DP-Pool in question. In monitoring the write-pending rate from the USP-VM on the CLPR exclusively used by the DP Pool in question, the spike immediately following dis-association/deletion of the V-VOL is quite apparent with write-pending rate jumping from 0% to ~50% for a period which would correlate with the time taken to zero-scrub the dirty pages of the V-VOLs being deleted.

My question is: Is such a performance impact to be expected when deleting DP V-VOLs on external disk, and if so, is there a way to limit this impact?