1

I am running Windows Server 2016 (RTM) in VMWare Fusion 8.5.1 on a late 2013 iMac (quad-core i5, 16 GB RAM) with macOS Sierra. At random moments the virtual machine practically grinds to an halt, usually corresponding with high CPU usage for the System process (50%, corresponding to one core). The machine is then extremely slow for several minutes, after which it usually passes (but sometimes it does not and a restart is required).

I am typically running Visual Studio 2015 with IIS Express on the machine while this happens, but I am unsure if there is a relation. Similarly, the issue seems to often appear after I have left the VM idle for a while.

Investigating the System process with Process Explorer shows that the culprit is ntoskrnl.exe!SeAccessCheckEx, which appears to be a routine related to UAC. However, disabling UAC does not help.

The VM has been assigned two cores and 8 GB of RAM.

Improve this question
2
  • The physical machine has 4 cores, I have assigned 2 to the VM (I have also experimented with different amounts of cores, to no effect).
    – Eepzy
    Commented Oct 31, 2016 at 18:26
  • Welcome to Super User! Instead of adding extra details that might help people answer the question in a comment, please edit the post and include the information there. That way, people are more likely to find it :)
    – Cas
    Commented Nov 8, 2016 at 14:24

1 Answer 1

Reset to default
1

I ran into this same issue today with a new install of Windows Server 2016 on ESXi 6.00 Build 3620759. Same thread ntoskrnl.exe!SeAccessCheckEx using 50% cpu (or 100% of one cpu depending on your point of view)

Numerous reboots did not fix the issue.

Guest was set to 6GB RAM and 1 virtual socket with 2 cores per socket.

I modified to 8GB RAM with 2 virtual sockets and 2 cores per socket, and the problem went away.

Changed the cpus back to 1 socket/2 cores per, problem still gone.

Changed RAM back to 6GB, Left CPU at 1socket/2cores per, problem comes back!

Changed RAM up to 8GB again, Left CPU at 1socket/2cores per, problem gone again.

So...it appears 6GB for this VM causes some issues. Unsure why.

Hope this helps someone.

Improve this answer
1
  • Thank you very much, your comment helped me a lot. I already had 8GB RAM and 1 virtual sockets and 2 cores. So I just changes number of cores from 2 to 6 without modifying RAM size and the problem went away. Used ESXi 6.5. It seems that just changing anything helps.
    – Maxim Masiutin
    Commented Apr 20, 2018 at 1:46

You must log in to answer this question.

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