Saturday, February 16, 2013

VOMA on ESXi 5.1 to find out Metadata Corruption on VMFS

I recently blogged about the how to verify the VMFS Heart beat corruption and in this blog I am going to show you how to use VOMA (vSphere On-disk Metadata Analyzer) to check if there is any inconsistencies after the events such as power outage or some others due to which everything went down at the same time.

voma can be used only on ESXi 5.1 and its not available in any prior version.

You can see the help of the voma command here.

Here are the screen shots of how voma will detect and identify the error found on the VMFS volume

As you can see in these example that there is file used against which the voma command was ran. e.g. .bin

Now before you run the command you may want to take a dump from each affected volume/s and then run the command against the dump file/s.


dd if=/vmfs/devices/disks/naa.0000000000000000000000000000:1 of=/tmp/naa.0000000000000000000000000000p1.dmp bs=1M count=1500

voma -m vmfs -f check -d /vmfs/devices/disks/naa.0000000000000000000000000000p1.dmp

Now as you can see the highlighted parts  (with gray background) in the above screen shots which are errors and inconsistencies found on the volumes. The white spaces are names which are hidden due to confidential information. If you find the errors on the volume/s then contact VMware Support for further actions.

There are five phases of the disk analysis and @VMwareStorage (Cormac Hogan) has posted about voma here and here.

You can check out the capabilities of voma on your own on ESXi 5.1.

Help by sharing !!

Thanks for your time.

1 comment:

  1. Great post!
    Discovered the value of voma today, the hard way, on ESXi 5.5, with a VMFS 5.60 filesystem on a >2TB volume: