ESXi

chain of events...

/user.b00 became corupt. ESXi would no longer load. (fatal error 33 inconsistant data)

Pulled drive, pluged into windows box, ran checkdisk, found and repaired errors.

Now hypervisor get most of way up, then purple screen of death.

tryed "recover" option, rolled back hypervisor sucsessfully to slightly previouse version, but same result when rebooted.

ran an install CD, it said "version that can be upgraded detected"... told it to upgrade...failed at 24% due to "directory not empty" It contained the "found" file from checkdisk, removed drive, deleted file.

ran an install CD, it said "version that can be upgraded detected"... told it to upgrade...failed at 24% due to "directory not empty" It contained the "recycler" file from windows, removed drive, deleted file.

ran an install CD, now it no longer recognises a version that can be upgraded, just that drive hase data. Stoped here.

I can access files on the disk via windows if I pull drive. Is their any way to extract the host config from this before I wipe it and re-install (no I did not backup host config agaid of time, just learned that I could have)

or, is there a way to force repair boot partition without loosing config?

Please help, any ideas would be great.

Save the file /bootbank/state.tgz to a safe location. Reinstall the host, then restore the configuration like described in VMware KB: Locally restoring an ESXi configuration from state.tgz backup

- Andreas

OK...ESXI makes multiple partitions, which one would contain this file? can't seem to find it. I have ESXi, Hypervisor1, Hypervisor2, Hypervisor3 partitions.

There should be two partitions that have a size of 250MB. They should both contain a state.tgz file. Pick the most recent one of the two copies.