I've had a Windows 2008 R2 server running fine in ESXi 4.1 for quite a while. Lately I added SQL Server 2008 R2 to the server and upped the RAM from 2GB to 4GB. I've also got two ubuntu 10.04 servers and a windows xp client running on the same host just fine.
Now I experience instability issues and strange file changes on the 2008 R2 vm:
1. A couple of times a week the server BSODs: The computer has rebooted from a bugcheck. The bugcheck was: 0x0000001a (0x0000000000041790, 0xfffffa800349b7b0, 0x000000000000ffff, 0x0000000000000000). A dump was saved in: C:\Windows\MEMORY.DMP. Report Id: 091911-19031-01.
Bugcheck 0x1a is a memory error.
2. I've got a small service that periodically scans all files in C:\Inetpub\ and compares the MD5 sums with the previous result stored in a SQL compact db (a sort of hacked-site-early-warning-system). About the same time as the BSODs started showing I also get alerts that files have changed. Comparing the file with previous backups confirms there are changes in perhaps a couple of bytes in the file. This causes random error on the web sites as DLL's, ASPX's and ASCX's change.
Often the change is reverted (and redeteced) after a couple of hours.
I've run chkdsk on the VM disks without it finding any errors.
Can there be issues with the RAM and or HDDs of the ESXi host?
1. Is there any memory error detection in ESXi host?
2. Is there any disk error detection in ESXi host?
I'm considering powering down the host, run memtest86+ then reformat and install ESXi5 instead.
I assume it boots the ESXi 4 VM's fine if I restore the VM's files?
