VMware

DRS and High CPU Usage on one host

We’re running 6.5 in our environment, 4 hosts in a cluster. DRS seems to like to keep host 1 at ~95-100% capacity on CPU, while the rest linger around 60-70%. Storage is shared on SAN, so only compute needs to move.

We can move VMs manually as much as we’d like, but DRS will typically move them back 10 minutes later, so I feel like this must be some kind of configuration issue. I recently changed the Threshold from 3 to 4, but that doesn’t seem to have made any difference.

At the moment, our hosts look like this:
Host1 – CPU 89%, Mem 45% (Master)
Host2 – CPU 72%, Mem 74%
Host3 – CPU 64%, Mem 63%
Host4 – CPU 72%, Mem 56%

Typically a server spiking is what would cause for the spike from 90% to 100%, but DRS just doesn’t do anything about it. We don’t actually seem to notice any performance issues when it happens, but it is pretty alarming to always see. Is there any reason that DRS seems to stack things like this? Would turning on Predictive DRS potentially help in this situation?

Thanks to anyone who can help!

Note: Resource Utilization is a bit higher than noraml barbecue we’re in between DR locations. We have a few more servers than we normally would.



View Reddit by b4mvView Source

 

To see the full content, share this page by clicking one of the buttons below

Related Articles

4 Comments

  1. I’ve had this issue before. We found that when we set DRS to manual with suggestions, we saw it wanted to move all VMs off of one host because of “resource requirements for VM not met” or something to that extent. We checked the power settings in BIOS and found that they were set to balanced. Changed them to high performance and DRS started moving VMs between all hosts as normal.

Leave a Reply