Jun 112014

If you are running VMware vSphere 5.5 Update 1 and are using NFS-connected storage, update your hosts ASAP (test first, of course). VMware just released a bug fix (express patch ESXi550-201406401-SG) for the All Path Down (APD) bug that has been plaguing NFS-mounted storage users for a couple of months now. Datastores can randomly go inaccessible for a few minutes, which is generally a bad thing.

If you find items in your log that resemble the following entry:

“Device or filesystem with identifier [12345678-qwerty0] has entered the All Paths Down Timeout state after being in the All Paths Down state for 140 seconds. I/Os will now be fast failed.”

…then patch quickly. NFS-mounted storage and heavy I/O workloads are prone to this bug. More details about the bug can be found at http://kb.vmware.com/kb/2077360/.

Update Manager has the patch available, or you can download the patch here.

I know that this was a tough bug to crack, and kudos goes to VMware engineering for tackling this one.