vPower NFS crashes while using to move large VM

[mbroeken]

Hi guys,

I'm currently using storage vMotion to move a big 5TB VM to my Veeam Box.
This VM is currently hosted on another NFS datastore (turned off Cloud Connect Replica).

I have many TBs left on my Veeam box and I was thinking on using vPower NFS to mount the nfs datastore to some of my ESX hosts. as Windows Server NFS is horrible, and there is no real alternative...
I know that vPower NFS is not that very efficient, but I did some tests and the performance was pretty good. This NFS datastore is not for production, but turned off Replica's. would be fine.

I went and tried to move the 5TB VM from one datastore to the vPowerNFS datastore. It all went fine, but around 6 hours later (around 2,5TB's). The vPower NFS process died.
I saw a lot of filesystem writes after the process died for around 5 minutes. Datastore was greyedout after that, I needed to restart the service to get it back running again.

Are there any tweaks to apply to make vPower NFS be more in sync with the Filesystem? (Like sync on unix, the number of threads? or any other tweaks?)
My best guess is that some buffer got too big and made the process crash.

Please advise,

[Gostev]

Hi, Marco - we will need to review debug logs for the NFS server and find the real cause of the process crash before making any recommendations. Thanks!