Unsolved

This post is more than 5 years old

13 Posts

416

October 7th, 2013 19:00

Migrating into FAST VP pool

IHAC that is migration  100+ TBs from various platforms into a VPLEX/VNX w/ FASTVP.  His experience so far has been that a significant portion of the initial allocations are on the NLSAS tier (they are a 4/16/80 mix)  This then causes an initial performance issue until things “balance” which is one to 2 days. ( a few were longer)  Their policy is “start high then auto” 

So…we are wondering if anyone has ideas or best practices that can help alleviate this issue…any thoughts will be appreciated.

13 Posts

October 7th, 2013 19:00

Most people so far that I've discussed this with suggest pinning to the low tier during migration.  Then only Hot data will be promoted during normal operations.

1 Rookie

 • 

86 Posts

November 21st, 2013 06:00

I would agree. Pin the low IO LUNs to NL-SAS and once things even out, maybe start switching things back to auto so the VNX can optimize the chunks of data appropriately.

No Events found!

Top