Hi all, I have two machines that are not identical but very similar: Phenom II processor, Gigabyte motherboard, 4-6GB RAM. Both are running CentOS 5.3 x64 and VMWare Server 2.0.1/2. Current kernel installed is 2.6.18-164.6.1.el5.
The web interface has become very unstable. It simply stops communicating at random times, usually when you do some change (change media on a virtual drive, change focus of one VM to another, etc) Consoles will close unexpectedly, hangs forever "loading" info into the various frames
Restarting vmware-mgmt services will usually bring things back. Sometimes it does not (blank page, logged in but "no access to console", starts up with with numerous "the server response included one or more errors" dialog boxes where the detail is "an object was not found")
Most disturbingly, the VM's themselves will sometimes just stop running. The guests are Windows Server 2003/2008 64 bit. One of the machines has a Windows 7 VM which oddly enough despite not being officially supported is the most stable. Running the other guests with the Windows 7 VM shut down doesn't change the behavior either.
These machines were much more stable before a massive yum upgrade of 150 or so packages. I have no idea which one it may have been, but I did boot up in the older kernels to see if there was any change, and there was not. I also tried 2.0.1 and 2.0.2 versions of Server with no change.
My real question isn't so much as "how do I fix this" but more of "where do I look to find out what's wrong"
1- How do I debug the web interface issues? (what are good logs to look at, etc)
2- How do I figure out why a VM suddenly terminates? (logs to look at there, etc)
Thanks!