Feature Request/Troubleshooting: Allow dismissal/Reworking of the maintenance modal in order to view alerts in maintenance mode

[NateTheSage]

As seen here, I experienced some kind of error in the database leaving Repomanager stuck in maintenance mode. While I could use some help on it, I'm more concerned with the apparent ability to see what the errors in the alerts below the modal, but the modal only offers the option to Refresh, which is not exactly useful if I can't get out of maintenance mode...

Would be happy to upload logs or something to the effect of helping sort the error, but as it stands right now the modal preventing me from working in the app is a bit of a pain, especially when the error description looks to be right there. Maybe making it a banner or something along those lines so we still have page navigation would be helpful.

Thanks for having made a great application!🥇

[lbr38]

Hello
You can try to delete the maintenance file by entering the container if you can, or from the host. This file is located in the data volume directory, for example on a standard install, the path from the host is /var/lib/docker/volumes/repomanager-data/maintenance.

Tell me if you are still blocked after that. Also send me the logs from the UI to see what caused this and fix it.
Thanks

[NateTheSage]

Managed to get in. My install is a little nonstandard since most of Repomanager's repo data is on an external drive. Nice to know about that maintenance lock file.

Seems my database is also locked.

SQLite3Exception: Unable to execute statement: database is locked in /var/www/repomanager/models/Stat.php:88
Stack trace:
#0 /var/www/repomanager/models/Stat.php(88): SQLite3Stmt->execute()
#1 /var/www/repomanager/controllers/Stat.php(106): Models\Stat->addAccessToQueue()
#2 /var/www/repomanager/controllers/Service/Unit/Statistic.php(222): Controllers\Stat->addAccessToQueue()
#3 /var/www/repomanager/controllers/Service/Execute.php(44): Controllers\Service\Unit\Statistic->parseLogs()
#4 /var/www/repomanager/tools/service.php(12): Controllers\Service\Execute->__construct()
#5 {main}

[lbr38]

It is not a problem if the Stats database is locked (this is temporary); it simply means that a write operation is currently in progress. The database can remain locked for several minutes (depending on the number of rows in the database) every Sunday around 01:00 for an automatic maintenance operation (cleanup, optimization, analysis). The error messages shown in your screenshot seem to correspond to this maintenance time window. Normally, all other write operations are paused while maintenance is running.

If this lock error occurs too frequently, then we will need to debug it to see what is blocking. Please wait until next Sunday to see whether new errors appear in the UI.

Regarding the maintenance mode that was not lifted, I think an error must have appeared in the container logs (docker logs repomanager), since maintenance mode is enabled when the container starts. I will look into adding more details to the logs to make debugging this type of issue easier in the future.

[lbr38]

In the meantime, you can also go to the /status page and check the latest database maintenance log to see if there was an error during the maintenance operation:

[NateTheSage]

It is not a problem if the Stats database is locked (this is temporary); it simply means that a write operation is currently in progress. The database can remain locked for several minutes (depending on the number of rows in the database) every Sunday around 01:00 for an automatic maintenance operation (cleanup, optimization, analysis). The error messages shown in your screenshot seem to correspond to this maintenance time window. Normally, all other write operations are paused while maintenance is running.

If this lock error occurs too frequently, then we will need to debug it to see what is blocking. Please wait until next Sunday to see whether new errors appear in the UI.

Regarding the maintenance mode that was not lifted, I think an error must have appeared in the container logs (docker logs repomanager), since maintenance mode is enabled when the container starts. I will look into adding more details to the logs to make debugging this type of issue easier in the future.

Alright, I'll see what happens around then. The stats endpoint must be a newer feature, how cool!

Looks green across the board thus far though.

Latest maintenance file I have is on the 8th, looks like it was fine.

[Sun Feb 8 01:00:00][INF] Enabling maintenance page
[Sun Feb 8 01:00:00][INF] Starting maintenance on database main
[Sun Feb 8 01:00:00][INF] Performing VACUUM on database main
[Sun Feb 8 01:00:02][INF] Performing ANALYZE on database main
[Sun Feb 8 01:00:02][INF] Performing integrity check on database main
[Sun Feb 8 01:00:03][INF] Maintenance completed successfully on database main
[Sun Feb 8 01:00:03][INF] Starting maintenance on database stats
[Sun Feb 8 01:00:03][INF] Performing VACUUM on database stats
[Sun Feb 8 01:14:52][INF] Performing ANALYZE on database stats
[Sun Feb 8 01:18:04][INF] Performing integrity check on database stats

[NateTheSage]

Come Sunday and it's so far so good. Wasn't sure what the problem was that occurred, but it seems to be alright now. Thanks for the support, though I do maintain there should be a less intrusive way to view alerts and not be stuck on a modal that has only the option to Refresh.

[lbr38]

Glad it works now!
I agree with you about the alerts, I will think about a better way to display them in the future, I'll keep this issue open for that.