vmware-vdiskmanager -R

[continuum]

A few days ago I had to fix a WS 11 vmdk on a Windows 8.1 host.
As I had expected the current version was not able to fix the problem.
The old binary from WS 7.1 complained about missing dlls - after we supplied the missing dlls it worked but in the end also failed to repair the vmdk.

So at the moment there is no tool available that can repair current vmdks.

Are there any plans to fix the broken vmware-vdiskmanager ?

Does any version of the Virtual Disk Development Kit has a working -R function ?

[dariusd]

Hi continuum,

Myself and one of the other engineers here recently took a few days to dig through the sparse disk checker code.  We ended up making quite a few improvements which should make it much more able to repair some of the more straightforward forms of corruption.  There will still be some disks that it's not smart enough to repair (the heuristics are not Artificial Intelligence), but you should see a huge improvement over the current versions.

Since I know you have a great degree of interest in it, I'll see if I can get you a pre-release build of the improved version of vmware-vdiskmanager to try out... We'd love to get your feedback on it.

Cheers,

--

Darius

[a_p_]

Just another EHR: It would be great if you could make the tool more "blathering", i.e. provide verbose output to either the console or a logfile even (or especially) if the tool cannot automatically fix the issue. So far I created my own tool (actually an Excel VBA module) that I use for basic sparse file checking, but it only provides a limeted output of what's wrong, and fixes need to be done manually. A detailed output/log for e.g. grain directory/table inconsistencies, size mismatch, holes, ... would be awesome.

André

[dariusd]

Hi André!

After running vmware-vdiskmanager -R, there should be logs in /tmp/vmware-${USER}/vmware-nnnnn.log and /tmp/vmware-${USER}/vdiskmanager.log – and in equivalent places on Fusion and on WS/Windows.  I do not know why there are two logs with much the same content...

Those logs usually receive a moderately-verbose listing of mismatches between the grain tables and how they are resolved, with the caveat that the dump won't be complete if we bail out (or fail hard) partway through.

Cheers,

--

Darius