2010-11-08 10:28 AM
I have a few questions revolving around the best way to gather usage statistics, specifically volume stats.
There is the obvious sysstat command but that is a very high level overview of the filer and what it is doing, I want to get more granular than that.
I've messed around a little bit with the "stats" command but it is very hard to use unless you build or use some of the preset counters, the man file is helpful but still vague at best.
Are there any easier ways or possibly anything I'm missing here that can assist?
What I am attempting to do is isolate a volume within an aggregate that is causing poor performance of the volumes it happens to share disks with.
2010-11-08 06:28 PM
I would speak to your Netapp representative about access to the CMPG website (cmpg.netapp.com), you can upload (or get them to upload one for you) performance stats.. The outputted spreadsheet is very very impressive and will outline lots of stuff, but what is pertinent for you is that it will outline the 5 top volume write, the top 5 volume reads and the top 5 volume latencies.....
Plus it will give you info that you never even thought the filer could gather.....