After attending Irfan Ahmad’s session on Storage IO Control at VMworld I had the pleasure to sit down with Irfan and discuss SIOC. Irfan was so kind to review my SIOC articles(1, 2) and we discussed a couple of other things as well. The discussion and the Storage IO Control session contained some real gems and before my brain resets itself I wanted to have these documented.
Storage IO Control Best Practices:
- Enable Storage IO Control on all datastores
- Avoid external access for SIOC enabled datastores
- To avoid any interference SIOC will stop throttling, more info here.
- When multiple datastores share the same set of spindles ensure all have SIOC enabled with comparable settings and all have sioc enabled.
- Change latency threshold based on used storage media type:
- For FC storage the recommended latency threshold is 20 – 30 MS
- For SAS storage the recommended latency threshold is 20 – 30 MS
- For SATA storage the recommended latency threshold is 30 – 50 MS
- For SSD storage the recommended latency threshold is 15 – 20 MS
- Define a limit per VM for IOPS to avoid a single VM flooding the array
- For instance limit the amount of IOPS per VM to a 1000
fletch00 says
is there any word on when SIOC will be available for NFS datastores?
AFAIK, the latency stats are now available for NFS, so it should be technically possible?
thanks
Tomi Hakala says
I got impression from VMworld last week that SIOC for NFS is coming in next vSphere release which is targeted to be released just before VMworld 2011. So next summer I’d guess.
Rob B says
Just a quick question on your second bullet-point above…
“Avoid external access for SIOC enabled datastores”
Does this include LAN-free backups with VCB / the new storage API’s? Or could you give some other examples of what external access would be in this context?
Alan says
i think external access means for example, esx 3.5 hosts, that doesn’t support SIOC..
Duncan Epping says
any type of node that would skew the average latency computation…. But with VCB you have a very good reason to not stick to this best practice of course. it is a best practice and not a requirement.
PiroNet says
@Rob B this includes SAN tooling such RAID config rebuilds or disk scrubbing for example.
Read more at http://kb.vmware.com/kb/1020651
Rgds,
Didier
Kuba says
Good post – I’ve only one comment – SIOC shouldn’t be enabled on datastores protected by SRM (http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2032365) as this would cause error during planned migration.
VMware Fanboy says
Here’s a nice article saying where abouts in the range you recommend to set the congestion threshold:
http://www.itsupportforum.net/topic/how-to-set-the-vmware-storage-io-control-congestion-threshold/