Hello
I encountered a very strange problem with a Windows 2003 VM and have no idea how to solve it.
The problem started when the VM was cold migrated from an old host to a new one, got a VMware Tools update and the requested reboot. Then the VM did not start anymore, resetting after less than a second (even before showing any Windows splash screen).
Even Safe Boot, Last Known Good etc. didn't really help; it would maybe start till the logon screen in safe mode but resetted (without blue-screen) a few seconds later.
I tried a lot of things: attaching to 2nd VM for chkdsk, disabling VMware Tools service by editing the registry hive, recovery console, new vm with just disk attached, etc. pp.
No joy. Then I configured the VM with just 1 CPU and 1 GB RAM, and it started! Tried 2nd CPU, 4 CPUs, more RAM - bang, problem back. I iteratively narrowed it down to this, 100% reproducible:
4032 MB RAM works
4036 MB RAM doesn't work :-O
To make it even sweeter, if I hot-add RAM (beyond 4032 MB) after the VM started, it works! To make it even more interesting, I can reboot the VM then and it still works! But if I shutdown the VM and restart the VM container with more than the magic 4032 MB, it resets immediately! :-O
I have another "identical" VM on that cluster, and there's no problem at all.
Any hints where to look?
More information:
- vHW9
- 4 vCPU, 8 GB vRAM (initially)
- The VM was originally (long time ago) created by P2V I guess (it has a virtual USB controller device)
- W2K3 Enterprise 32 bit SP2
- Memory Hot Add enabled
- ESXi 5.1.0 build 1483097
- Critial system files are identical to a working system (hal.dll, ntdetect.com, ntldr, ntoskrnl.exe, boot.ini)
- chkdsk, fixmbr, fixboot done
- VMware Tools reinstalled
- Booted on different hosts (even on the "old" cluster)