VMware Cloud Community
hamel01
Contributor
Contributor

Disk Latency spikes

Hey Everyone,

I am looking for some input on a odd issue i am seeing.  I have a new HP DL380 Gen9 with (8) 600gb 15k sas drives in raid 10.  Esxi version is the latest 6.5 build for HP and the server is patches with the latest SPP from HP.  The server has dual CPU's and 196gb ram.  Currently it is running (3) freshly built 2012 R2 servers and (3) P2V's 2012 R2 servers as well as (1) 2008 R2 P2V'd server.  Everything seems to work fine but when I look at the monitor tab for the host and change the view to Disk I get the graph that is displayed below.  I am trying to figure out why these latency spikes are occurring.  I have powered off all VM's besides one of the DC's and the spikes are still there.  Time of day doesn't seem to matter as well.  I plan to do my test again this weekend and shut down the remaining DC as well but i doubt that is going to show me anything different.

pastedImage_0.png

0 Kudos
1 Reply
daphnissov
Immortal
Immortal

It may likely be RAID checking or SMART validation from the controller itself which is temporarily increasing latency to the back-end. You may be able to correlate the spikes with log entries around the same time to see if there are SCSI sense codes being returned.

0 Kudos