2010-03-30 05:05 PM - edited 2015-12-18 01:35 AM
We have SQL 2005 (soon SQL 2008) on a FAS3160 unit, 28 spindles, FC SAN, 15K disks. This is the only host using this storage system. We're probably at 60% of total space capacity.
It seems like since we went live 6 months ago on this storage system (NetApp is new to us), that performance has gradually degraded. SQL Server is increasingly reporting waits from the disk subsystem.
Is there any checklist or guide for how to keep your NetApp box running at peak performance, in a SQL Server environment? We do all of the typical SQL Server DBA maintenance (index rebuilds and so on).
2010-04-03 09:32 PM
Do you have Performance advisor in your environment? Are your 28 spindles in a single aggregate or are 14 disks per head? If you've split them in half, then you should have 2x 11 data + 2 Parity + 1 spare?
2010-04-04 01:25 PM
I'll have to ask the storage team for specifics (I'm more of a SQL guy than storage), but I believe Yes we have performance advisor, and we have a single aggregate.
We did read and follow the best practices for SQL on Netapp (the other doc posted), but that's more of a setup and sizing guide. I'm wondering more about ongoing care-and-feeding of NetApp specifically in a SQL environment.