Hi All,
We have a NetApp with two NFS datastores acting as the SAN for our 5 ESXi servers (all running ESXi, 5.0.0, 1117897). We are using View 5.2.0 build-987719 . We are using linked clone Windows 7 floating desktops and we have 2 intermittent problems.
1. Sometimes when we build a pool we will see this error inside of vCenter
--------------------------------------------------------------------------------
Clone virtual machine
replica-13dd0c3d-98e4-48cf-9480-6e654aa0e437
"The virtual disk is either corrupted or not a supported format"
--------------------------------------------------------------------------------
The pool seems to complete fine
2. The more significant problem is sometimes when we recompose a pool, delete a number of desktops in a pool, or delete a pool outright we will see a significant number of errors like this
---------------------------------------------------------------------------------
Delete virtual machine Win7PUBTEST-28
"A general system error occurred: Input/output error"
---------------------------------------------------------------------------------
When this happens (during a pool deletion process) the pool no longer appears inside of VMware Horizon View Administrator but all the desktop folders that had the above error still exist in the NFS datastore. Event if we try and manually delete the orphaned desktops in the NFS datastore it won't let us, sometimes we have to wait 15 minutes, other times it can take up to an hour. Some process seems to be locking the vmdk files inside the desktop folders.
We are have been using VAAI but have also tried building pools without it and instead using the built in View Storage Accelerator but we've had issues with both.
Thanks for any advice you can offer.