Request to Include SQLite in plex-tmpfs for WAL Management

[silkyclouds]

Hello,

I’ve been testing controlol/plex-tmpfs on Unraid 7.0.0 with a large Plex library (~8 GB SQLite DB). Moving the DB into a 32 GB tmpfs has been promising, but I ran into two issues:

1.	WAL files grow during scans and don’t shrink automatically, eventually causing “database or disk is full” errors.
2.	The stock image doesn’t include sqlite3, so I had to deploy an external “wal-keeper” (Alpine container sharing IPC/PID with Plex) to run PRAGMA wal_checkpoint(TRUNCATE) every 15 min and VACUUM every two days.

Managing WAL manually is cumbersome. Would it be possible to bundle sqlite3 directly into controlol/plex-tmpfs? That way, users with large libraries could add a simple in-container cron or s6 service to checkpoint and vacuum the DB without an extra container.

Thanks for considering this enhancement—any feedback or alternative suggestions would be greatly appreciated.

[controlol]

That sounds like a good idea! Haven't run into this issue yet personally.

To be honest I expected Plex or SQLite to manage these kind of things, this image was not intended to edit the actual sqlite files.
Can you check if wal_autocheckpoint is already enabled? https://www.sqlite.org/pragma.html