Disk Performance high?

[escapem2]

hi guys

ESX 3.5 update 5, 6 ESX host

Yesterday we got an issue where 15 VMs were completely frozen and the rest another 30 were working pretty slow.

That was around 2:30 yesterday today I checked the disk statistics per ESX host and I found this check my attachment

Is that Disk referring to the local ESX hard drives or storage? either way what should I check? What could be causing this issue? Could this Hard drive behavior caused my system to crash?

Thanks a lot

[PaulSvirin]

What was the network utilization directly at that time?

Don't you have the snapshotting routines croned at that time?

You can wait for next day to check whether this occurs at the same time, this can be a hint: http://communities.vmware.com/docs/DOC-5490

---

iSCSI SAN software

http://www.starwindsoftware.com

[escapem2]

adding CPU which looks normal for me and memory and networking take a look thanks a lot

crontab -l shows nothing, vmkernel no info at all, something else I could check ?

about storage stuff

I already check DAVG - KAVG - GAVG I should say sometimes the DAVG value is above 30 sometimes but it is for a second just that I am trying to understand this esxplot tool http://www.yellow-bricks.com/esxtop/ since I see to many Physical disk parameters

I really don't know when I should be worried about DAVG I mean if the value is steady for 10 secs I should know I got a problem? or the value all the time should less that 25 secs?

thank a lot

[RParker]

15Mbs throughput isn't high, but it depends on the RAID and how you have it configured. How many disks, type of RAID, and the type of drives as well.

If this is a SATA RAID 5 with like 3 7200 disks, you are done. Not much tweaking you can do. If however you are looking at 15K SAS/SCSI drives on a battery backed RAID controller with 6 or 8 drives, that shouldn't have that much impact with that low of disk, but then we don't "see" the IOPS either...