0

After a (patch and) reboot, I got an error message reading something about corrupt recycle-bin for e:\ (where I store my virtual machines). Two of the virtual machines could not be started. On the first of them, it seems permissions had been changed - after using icacls I got it up and running. The second however has a missing VHD disk.. It has two, one for system and one for data, and the one for data is 0 bytes on the filesystem.

I assume the data is gone on the (physical) disks, and I will have to revert to last backup - but this kind of scares me so how can I 1) see exactly what might have happened and 2) see if the E:\ disk is actually 100% healthy?

Any help is much appreciated

3molo
  • 4,330
  • 5
  • 32
  • 46

1 Answers1

1

I'm going to get to your answer, but first I want to make sure you do 2 things in the future:

  1. Use RAID. It's not a backup, but the redundancy helps. This could have been caused by a bad hard drive or bad sectors on the drive. A RAID array would have avoided this downtime
  2. Use and UPS. If the power gets cut on the server at an inopportune time, filesystem corruption can occur

Now for your answer. Run the following command to ensure you've got a healthy filesystem and no bad sectors. This has the advantage of tackling both parts of your question. Errors in the first stages indicate filesystem damage. This is most often caused by improper shutdowns. It can also be hardware related. Errors in drive scan stage indicates bad sectors on the drive. I'd replace the drive if you find a lot of these.

chkdsk /r /f e:

Jason Berg
  • 19,084
  • 6
  • 40
  • 55
  • It's a DAS, with RAID 10. We do have multiple UPSes. Tell me, is the chkdsk read-only - if not, I wouldnt dare to run it without first migrating all virtual machines.. – 3molo Oct 13 '10 at 07:19
  • With those switches, it is not a read-only process. – Jason Berg Oct 13 '10 at 15:33