1

We have a Dell PowerEdge R820 running VMware ESXi that has had multiple CPU errors over the past two months. It was originally having an IERR on CPU1 and now is showing CPU Check errors for CPU3. I swapped out CPU3 --> CPU2 and CPU1 --> CPU4 to see if the errors would follow the processors around. After swapping, the server was on for about 72 hours before I performed another DSET and there were no more hardware errors. I want to run a stress test on the CPUs and memory for about 24 hours to verify it will be able to handle the load when we go into production. It's possible the CPUs were somehow shifted during shipping (soon before we began receiving the first errors).

The problem is that I don't know of any stress test tools for my particular setup. I tried using breakin but it won't boot up into it correctly. (Gets hung up on 'loading sb_edac'). Is there a tool that you can recommend I use to stress test an R820 (Intel Xeon e5-4600) running VMware ESXi?

1 Answer 1

0

In this case, you can just load a Linux live-CD or rescue CD containing the stress tool (I prefer the PLD Linux disc) and leave the server on a loop.

However, I would lean heavily on Dell support and replace the system board/CPUs and/or RAM. The server is not healthy! Debugging this type of thing may not be time-effective.

Did you capture the dump of the ESXi system? Support logs?

2
  • I have found Ultimate Boot CD which contains the stress tool. I can try that next. I have leaned on Dell support and, after recommending I switch out the CPUs, they recommended running a stress-test for 24 hours. I collected DSET logs for before I swapped the CPUs out and three days after I swapped them out. I also have a dump of the VMSupport logs.
    – Robert Ray
    Sep 17, 2013 at 13:49
  • 1
    Well, go ahead and run the tool. Send the VMware support bundle to VMware.
    – ewwhite
    Sep 17, 2013 at 14:07

Not the answer you're looking for? Browse other questions tagged .