Lost NFS after renaming old & presenting new with the old name

The host is losing access to the existing NFS share when the user renames the share and present a new share with the old name of an existing share.

Steps to Repro:

1. Map an NFS share to host with any name, for example, “NFS-01.”

2. “NFS-01” has a shared path as “/nas/nfs-1.”

3. Rename “NFS-01” to “NFS-01-renamed”

4. Present a new NFS share to the host with the share path “/nas/nfs-2.”

5. Map new NFS with the name “NFS-01”

Error:

The "Create NAS datastore" operation failed for the 2 entities with
the following error message.

An error occurred during host configuration
Operation failed, diagnostics report: Unable to get console path for
mount

Why Host loses access to the NFS share :

In current versions host uses volume labels, i.e. “NFS-01” as part of the device name. When NFS datastore was created with the name “NFS-01”, the device name was “nfs:nfs-01”. When a user renames NFS share name, the host will not update the device name, it remains “nfs:nfs-01”.

When a new share gets presented to host with “NFS-01”, it also creates a device with the name “nfs:nfs01” which is when the host sees a conflict. Which is causing the issue.

Workaround:

Rebooting the hosts where these NFS shares were presented

Fix :

The issue is fixed in 6.7GA. It will also be backported to older versions soon.

Reference:

https://kb.vmware.com/kb/56577

Leave a comment