I've been running Exchange 2003 virtualized one Server1 almost since Server1 was in general release. Exchange size is similar to yours, around a dozen mailboxes. Host is much older, an old Compaq Proliant, PIII/733MHz, 1.25GB RAM. SCSI HDs instead of SATA. Win2k3 host OS. Currently on Server 1.0.3 because I haven't had time to update it to 1.0.4. Performance is a tad slow only because the host is slow, but otherwise, I've had no problems with Exchange VM on this host. Host also runs one other Win2k VM.
While it could be an issue with v1.0.4, I'm not inclined to believe so unless someone has other evidence. This was going to be a production server? You do realized that XP is not a supported host OS for Server, although there really isn't a problem running it this way IMO. But no mirrored HD on the host? Like MS is saying, are you sure it's not hardware? Have you checked the host's HD?
Edit: Although you say this is a new machine, you mentioned that 1GB of RAM "failed." That's not giving me assurance on your hardware. 
Thanks for the reply.
I was on server v1.0.4 but am currently running Workstation v6... I think I will stay with Workstation as the background snapshots and the snapshot manager are tools useful to me recently given the issues I have been having, it's been a good piece of mind taking snapshots at given points.
WS6 is fine. I prefer Server for a production situation like this because Server can autostart a VM as a service. No manually starting the VM or creating a script to start the VM. As for the snapshots, I went "low-tech." Shut down VM and copy the entire VM directory. Yes, it would be nice to have multiple snapshots in Server, but there are issues with trying to move a VM to another host if you have snapshots. So for simplicity, I don't use snapshots much.
The ram was non-hp that I brought and a stick failed, the machine had Vista on it and the failed ram caused issues with the stability of the OS and so I removed it and installed XP and it's been running fine since.
Ok
Other then running a chkdsk, are there any other tools that would check the host hardware out more fully ? for what it's worth the host's event logs are clean...
Well, that'd be a place to start. Rule out the obvious. Make sure it's not hardware like broken fan or something. If your Exchange store is already so corrupt that you can't repair, and possibly the corruption had been happening for a while so your backups are also corrupt, how about just starting with a new information store. Yes, a pain to recreate everything, but with only a few mailboxes, I'd try to get everyone's mailbox backed up to a PST, then reload the PSTs when you create the new information store. (I've done that in the past as it was a known amount of time to get everything back running rather than keep trying to fix the store and praying that it's still not corrupt.)
Also, since you changed from Server to Workstation and you still get corruption, how about installing Server or Workstation on another machine to see if you still get corruption. For so few mailboxes, you don't need a lot of horsepower on your host, as my setup illustrates. 
Thanks again
edit: Reason for not mirroring the HDD's is that I figured with regular backups of the Virtual Machines directory I could just use another machine if this failed (go buy one and install VM on it), mirrored drives would not increase performance/read write speeds though would it....
Well, this is how you test your backup, I guess. :smileysilly: Without RAID or at least mirrored HDs, you are more dependant on your backups. And how much downtime are you willing to accept while you move the VM to the new HD or new machine. Also like I said above, snapshots can cause issue when moving the VM to a new machine. (Specifically absolute paths. You have to configure the new machine with the same path to the VMs.) In a production environment, I never have anything less than RAID1. "Cheap insurance vs. cheap"is how I sell it to the clients. Development machines I'll live with no RAID, but not a production machine. This is not about performace. This is about reliability.
And if you didn't notice the corruption right away, how long do you archive your backups? (Since you're more reliant on them without RAID1.) To be blunt, here you have a good example of how your DR procedures need improving. If your Exchange is not working now, your users aren't screaming at you? My clients are mostly small offices usually with less than a dozen people each. All of the ones using Exchange or other internal mail or collaboration app would be coming after me with pitchforks if the server is down for more than a day. 
(Sorry to be preaching on this last point, but I am always having to sit the client down and discuess the dangers and ramifications when the client doesn't think reliability and thinks cost. Or isn't paranoid enough about thier data. As the saying goes, there are two kinds of people: those that have lost data, and those that will lose data. I've had clients redo a month's worth of work because they thought that they had good backups but later found out that the person who was checking the logs or changing the tapes...wasn't, yada, yada, yada. Ok, I'm off the soapbox now.)
Perhaps someone else has other ideas....
Thanks for the reply - the data is as below:
I dont fully understand the data so would this suggest that the disk should be replaced ?
thanks again