Feature request: Automatic unpublish of NFS datastore

[cdesch]

Hi all,

when running a Linux FLR, the Helper VM gets deployed via instant recovery. Due to IR a vPower NFS Datastore gets mounted on the ESXi host, choosen in the FLR wizard.
But when the FLR recovery is done -> nobody cares about "unmounting" the vPower NFS datastore from ESXi.

When recovering IR with a full VM Image, after Storage vMotion -> Veeam automatically unpublishes the temp NFS Datastore, so why it is not done during Linux FLR ?

Problem: ESXi is polling connected Datastores, NFS datastores too, an if the vPower NFS Proxy i.e. needs an maintenance & reboot, all ESXi Hosts, that ever ran a Linux FLR, has stale NFS mounts.
ESXi is not happy seeing datastores which are not accessable.

Please give us an RegKey or another functionality to control the "auto unpublish" of an vPower NFS temp datastore.

Thx, Chris

[veremin]

We've got similar requests already. So, thank you for raising this.

As to a workaround, you can create a vCenter alarm using VM removal event as its trigger. Leverage a condition that alarm should be only triggered if the datastore name starts with "VeeamBackup". Finally, set a PowerShell script (that unmounts the datastore) as alarm post-action.

[foggy]

As far as I can recall, the datastore is left mounted intentionally, in order to save time on mounting it again during the next vPower task.

[Cragdoo]

I just came across this same issue today. Setup and ran a one-of Surebackup job 2-3 weeks ago and the NFS datastore was still presented to the cluster. Some kind of timeout then NFS unmount would be nice. ...sorry for resurrecting an old FR

[nielsengelen]

For those who want to auto unmount, the following PowerShell script can be used via Windows Task Manager to unmount the vPower NFS datastore. Just run it every x hours and it will do it’s job.

Script is available via GitHub: https://github.com/nielsengelen/veeam-p ... astore.ps1