I bought and installed Workstation 8 as an upgrade to my 7.1 license, and things got so bad that I uninstalled it and reverted to 7.1.
Once started, my WinXP guest would initially seem to freeze for seconds at a time. These pauses grew longer and eventually worked their way into the host OS. 30-45 minutes after starting Workstation 8 and guest, my entire host essentially became unusable for minutes at a time with short reprieves of 10-20 seconds between these freezes, so I reverted to 7.1 and have been fine ever since.
This happened whether I ran the original v7 guest VM or the upgraded v8 guest VM.
Has anyone on Maverick experienced this? uname -srvmo returns:
Linux 2.6.35-30-generic #59-Ubuntu SMP Tue Aug 30 19:00:03 UTC 2011 x86_64 GNU/Linux
Unfortunately more people are experiencing guests freezing up:
http://communities.vmware.com/message/1833469
No word from VMWare so far ...
The odd guest hiccup I could possibly handle, but when it affects the HOST and grinds everything else to a halt, well that's when it crosses from bad to unacceptable.
I have no problem with WS8 under a Windows 7 host, just Ubuntu. Very disappointing.
--I was seeing this issue with the Beta, but not with "gold code" WS 8 release. How much RAM do you have on the host, and how many vCPUs and RAM are you giving the VM ?
8GB RAM, quad core Intel, running just the single 2 vCPU VM.
--Try decreasing the vCPUs in-VM to (1) and monitor disk I/O with ' iostat -k 5 '
and RAM/CPU usage with ' top -d 5 '
What happens when you power-on the VM ? Does host start using Swap and CPU ?
--Best practice is to have all VMs on a separate disk-spindle from the Host OS; are you overcommitting RAM for the guest ?
High (insanely so) disk access seems to be the problem. I reinstalled 8, and tried moving the VM to a USB drive. It did make things better, but it's much slower than it was in 7.1.4 where I had no problems running against the internal disk.
So basically, what the heck changed between 7.1.4 and 8.0.0 to cause disk access to go wild like this?
FWIW, changing the vCPU count to one did nothing, and the VM only has 1GB RAM allocated to it which is a fraction of my physical RAM, so it's not a swap issue.
I just reported a somewhat similar issue with a very similar hardware/software setup: