Change blob store to NFS share is failing until Stalwart-mail service restarted

[DKirk-ADS]

Debian 12 installation, been struggling for a few days to change the BLOB store to a NFS share on the NAS we have. I have the issues better documented on Reddit

https://www.reddit.com/r/stalwartlabs/comments/1df4rsy/change_blob_store_to_nfs_share_is_failing/

In short, once the change is made to the store the Stalwart UI breaks. Restarting the service manually seems to restore functionality again and the blog store on the NFS is working. Same with a server reboot, the UI faiIs and I must manually restart the service before the UI runs again. FSTAB is told to wait for the network before establishing the NFS share, it seems to be up after the reboot, but the Stalwart UI fails until the service is restarted.

Anybody have any thoughts on this? The NAS has a ton of storage space for the emails while the server disk space is very limited, hence my need to use the NAS.

[mdecimus]

In short, once the change is made to the store the Stalwart UI breaks.

This is because Stalwart stores the webadmin bundle in the blob store.

Anybody have any thoughts on this? The NAS has a ton of storage space for the emails while the server disk space is very limited, hence my need to use the NAS.

The reason why your blob store is not working is due to insufficient permissions:

Build error for key "store.synology.path": Failed to create database directory /mnt/synology.email/data: Os { code: 13, kind: PermissionDenied, message: "Permissiion denied" } 

Stalwart does not have enough permissions to create a subdirectory under /mnt/synology.email/data.

[DKirk-ADS]

Permissions are now fine, it created all sorts of files in the \mnt\synology.email\data directory, see below. The real problem I am having now is upon reboot the stalwart service needs to be restarted before the UI will respond, and once restarted everything works fine. The server doesn't have enough drive space for the 70+ users I hope to add, so I'm using the NAS storage. After changing the blob location it would be nice if the service would restart itself instead of failing.

After reboot this is the log entry, and simply restarting the service cures it. The service starting before the network/NFS?

2024-06-19T19:15:01.898222Z INFO common::manager::boot: Starting Stalwart Mail Server v0.8.1...
2024-06-19T19:15:02.331871Z ERROR smtp::queue::spool: Failed to read from store: Internal Error: No store configured context="queue" event="error"
2024-06-19T19:15:02.331961Z ERROR smtp::reporting::scheduler: Failed to read from store: Internal Error: No store configured context="queue" event="error"
2024-06-19T19:15:02.373904Z WARN jmap: Failed to unpack webadmin bundle. event="error" error=InternalError("No store configured")
2

Files in the blog "data" folder:

root@stalwart:/mnt/synology.email/data# ls -l
total 6772
-rwxr-xr-x 1 stalwart-mail stalwart-mail 1715218 Jun 19 14:08 000010.blob
-rwxr-xr-x 1 stalwart-mail stalwart-mail 1049 Jun 19 14:08 000022.sst
-rwxr-xr-x 1 stalwart-mail stalwart-mail 120068 Jun 19 14:08 000097.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 1232 Jun 19 14:08 000120.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 1233 Jun 19 14:15 000131.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 1173 Jun 19 14:47 000164.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 1172 Jun 19 14:47 000165.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 1128 Jun 19 15:17 000172.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 1103 Jun 19 15:17 000173.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 1379 Jun 20 11:30 000178.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 1572 Jun 20 11:30 000179.sst
-rw-r--r-- 1 stalwart-mail stalwart-mail 5080 Jun 20 13:08 000180.log
-rw-r--r-- 1 stalwart-mail stalwart-mail 16 Jun 20 11:30 CURRENT
-rwxr-xr-x 1 stalwart-mail stalwart-mail 36 Jun 19 14:08 IDENTITY
-rwxr-xr-x 1 stalwart-mail stalwart-mail 0 Jun 19 14:08 LOCK
-rw-r--r-- 1 stalwart-mail stalwart-mail 340674 Jun 20 13:41 LOG
-rw-r--r-- 1 stalwart-mail stalwart-mail 2268787 Jun 19 14:08 LOG.old.1718820920536988
-rw-r--r-- 1 stalwart-mail stalwart-mail 179468 Jun 19 14:15 LOG.old.1718821034985017
-rw-r--r-- 1 stalwart-mail stalwart-mail 167240 Jun 19 14:18 LOG.old.1718821249047793
-rw-r--r-- 1 stalwart-mail stalwart-mail 222573 Jun 19 14:38 LOG.old.1718822392460482
-rw-r--r-- 1 stalwart-mail stalwart-mail 167278 Jun 19 14:42 LOG.old.1718822847190922
-rw-r--r-- 1 stalwart-mail stalwart-mail 177784 Jun 19 14:52 LOG.old.1718823832292654
-rw-r--r-- 1 stalwart-mail stalwart-mail 163468 Jun 19 15:13 LOG.old.1718824641843450
-rw-r--r-- 1 stalwart-mail stalwart-mail 1133283 Jun 20 11:30 LOG.old.1718897443447872
-rw-r--r-- 1 stalwart-mail stalwart-mail 3122 Jun 20 11:30 MANIFEST-000181
-rw-r--r-- 1 stalwart-mail stalwart-mail 102099 Jun 19 15:17 OPTIONS-000177
-rw-r--r-- 1 stalwart-mail stalwart-mail 102099 Jun 20 11:30 OPTIONS-000183

[mdecimus]

After reboot this is the log entry, and simply restarting the service cures it. The service starting before the network/NFS?

Probably that is the problem. You need to change the service file so Stalwart starts after the NFS volume has been mounted.