• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

ESX 4.0 server - cpu usage randomly drops

R

Red Squirrel

[H]F Junkie
Joined
Nov 29, 2009
Messages
9,217
Edit: It's actually ESX 4.0, my bad.

We noticed when looking at the graphs that sometimes the cpu usage will just dive for a few seconds, where a host is using 0%. Today I managed to find a pattern, don't know if it's always been this pattern or not, but I happened to be working one of the VMs and it just locked up.

It seems it will do it for about 10-15 seconds, at which point all VMs on that host are inaccessible / not operating, then it will be ok for about 4 minutes, then it will dive again for 10-15 seconds, then it will be ok for about 25 minutes, then the cycle restarts. When this happens, it also happens to another of our hosts, at the same time. We have 4 hosts in that cluster, the other two seem fine.

Anyone know what's going on?

We have an Equalogic iSCSI SAN as the storage back end. If you need more info I can provide it.
 
Do you happen to have any VM's that have multiple vcpu's assigned to them that are doing heavy cpu intensive tasks?
 
Does not look like it, and the fact that it does it to two hosts at once seems kinda odd. But no does not look like anything is out of ordinary, there is a vm with 2 cpus. Checking the rest now.
 
Turns out it was vranger that was causing this. We'll have to find a better solution.
 
Could you give some details about this? Like how it was setup that was causing this?

What are you looking at to replace vranger?
 
I don't know the exact details, my coworker is prime on this, but all I know is that when he stopped vranger, it stopped doing the blips. This was setup in a rushed fashion originally and it was setup wrong in so many ways.

It looks like Backupexec might have a module for backing up VMs but we did not look at it too deep yet.
 
Depending on the backup method(VADP vs VCB) you need to look at how many concurrent backups are being run per datastore(ideal 1).

If this is affecting a specific VM you may want to investigate the VMWare tools install/snapshot provider.
 
EQL array? Did you follow the dell doc (that has 3-1 or 2-1 vmkernel to pnic ratios)?
 
I did not set it up so I can't be of much help but yeah pretty sure all was followed. Though, I think we did have more then 1 concurrent backup. Maybe that was part of the problem. I'll have to check tomorrow.
 
If you followed that doc, and are on 4.0, you may have it configured wrong. Let me know.
 
You must log in or register to reply here.
Back
Top