I have a few servers on my network that are experiencing high CPU usage. It's nothing specific but everything seems to be running hotter. Task Manager is running at up to 6 or 7% instead of below 1. Same with RPC and a few other normal idle processes. These machines are idling at around 10-15% CPU usage while others without the issue are idling at 0-3%. It's already very odd that only some machines are having this issue, they are all configured the same and on the same domain following the same GPOs. There is nothing unique about them either, they are all development VMs but some dev VMs are fine. It becomes a huge problem whenever any intensive application is started such as Visual Studio, these machines shoot to 100% usage and become unusable. A normal dev VM runs VS with CPU at around 30-40%.

So I've been playing around with this for a few weeks and I found a band aid. If I change the machine name and the static IP, reboot, then change it back, it runs great. But only for a day. The next morning everything is back the way it was unless I do this again but obviously this isn't a viable long term fix. Any idea what on the system could be causing slow performance across the board but is magically fixed by cycling the machine name and IP? I'm baffled.

  • Have you tried running Sys Internals - Process Explorer? This program will tell you what in particular is using up the extra CPU. If you need to dig deeper you can use Process Monitor to dump a log of everything that is happening when the machine is thrashing at 100% CPU. This should give you a clear indicator of the root cause. – Lumberjack May 29 '18 at 14:40
  • "they are all development VMs but some dev VMs are fine" - the problem may not be your individual VM's but the cumulative load on the hypervisors they run on. Powercycling the VM's may result in them getting restarted on different less overloaded hypervisors. Do you have access to the hypervisors? – HBruijn May 29 '18 at 14:56
  • Thanks for the advice, procmon wasn't the biggest help since it seemed all the idle processes were just running hotter, it wasn't a specific process that was the issue and I didn't get any real new info from it as far as arguments or where it was initialized that helped me track down the issue. Also, we don't have access to the backend at all, although a ticket was in with them to see if they could check in on it for us. – Tony May 31 '18 at 10:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.