Frigate NFS mount point trouble

Configuration
1

I had frigate writing to an NFS share on my network, and today a power failure knocked everything offline. When the system came back up, frigate will no longer write to the NFS share, and instead re-creates a new set of files in the mount point folder on the HA host. I’m using Home Assistant 2023.6.2
I can see the NFS share advertised via another host on the network via showmount.

Is there no way to delete the frigate folder from the media folder? I couldn’t see a way to do it in the UI. I used the samba add-on to remove all the files from the mount point folder before trying to re-mount the NFS share from my storage machine. The NFS share seems to mount using the Storage function/feature of HA, but frigate chooses to write the files locally.

Has anyone got a solution for this?

2

a little more information… I have a raspberry pi 4 as my cam storage machine. it has a USB3 attached md raid5 disk. I made an NFS share at /store1/frigate. Everything was working before the power failure.

root@camstore:~# showmount -e
Export list for camstore:
/store1/frigate 192.168.1.66,192.168.1.84
root@camstore:~# exportfs
/store1/frigate
		192.168.1.84
/store1/frigate
		192.168.1.66
root@camstore:~# ll /store1/frigate/
total 2208
drwxr-xr-x 2 nobody nogroup    4096 Jun 22 13:43 clips
-rw-r--r-- 1 nobody nogroup   57344 Jun 21 22:22 frigate.db
-rw-r--r-- 1 nobody nogroup   32768 Jun 22 13:44 frigate.db-shm
-rw-r--r-- 1 nobody nogroup 2154792 Jun 22 13:44 frigate.db-wal
drwxr-xr-x 3 nobody nogroup    4096 Jun 22 06:09 recordings
root@camstore:~# cat /etc/exports|grep -i frigate
/store1/frigate		192.168.1.84(rw,sync,no_subtree_check) 192.168.1.66(rw,insecure,all_squash,no_subtree_check,async)
3

… and just like that. Last night brought another storm with a power outage. When I brought HA up this time it mounted the NFS share from 11 days ago. I tried rebooting several times earlier in the past couple of weeks trying to get frigate to use the share.

There have been a couple of updates since the original problem, but I didn’t recall seeing anything related to my situation. I’m glad it’s back, but I worry for how long?