I found it by becoming properly annoyed and writing a DB -vs- DB direct comparison tool. That’s something I should have had from the beginning but … meh lol.
FYI… looks like 1.43.4.10903 has the schema changes for indexing as there is a bunch of Completed forward migration 202608120900. along with [FullTextSearch] Recreated the search tables, repopulating them., Analyzing database. and Vacuuming database. upon the initial startup after upgrading.
Initial server startup took a little longer than normal on my pretty powerful system (20 core Ultra 7 265K) so it might take longer on slower systems, just let plex do it thing …
Yes, I’m looking at that. I’ve had 1.43.4 in my hands since first built.
I am planning around the transition . (Plex takes over the FullTextSearch tables)
BTW… I’m still using version v1.15.00 of your tool and for the first time since the checks for the fts indexes was added they now pass “ok”, even after multiple library scans.
[2026-08-25 18.58.34] Stopping PMS. (60 second max delay)
[2026-08-25 18.58.34] Stopped PMS.
[2026-08-25 18.58.34] Checking the PMS databases
[2026-08-25 18.58.45] Check complete. PMS main database is OK.
[2026-08-25 18.58.59] Check complete. PMS blobs database is OK.
[2026-08-25 18.58.59]
[2026-08-25 18.58.59] Checking FTS (Full-Text Search) indexes
[2026-08-25 18.58.59] FTS index 'fts4_metadata_titles' - OK
[2026-08-25 18.59.01] FTS index 'fts4_metadata_titles_icu' - OK
[2026-08-25 18.59.01] FTS index 'fts4_tag_titles' - OK
[2026-08-25 18.59.01] FTS index 'fts4_tag_titles_icu' - OK
[2026-08-25 18.59.01] FTS index 'fts4_metadata_titles' (blobs) - OK
[2026-08-25 18.59.01] FTS index 'fts4_metadata_titles_icu' (blobs) - OK
[2026-08-25 18.59.01] FTS index 'fts4_tag_titles' (blobs) - OK
[2026-08-25 18.59.01] FTS index 'fts4_tag_titles_icu' (blobs) - OK
[2026-08-25 18.59.01] FTS integrity check complete. All FTS indexes OK.
[2026-08-25 18.59.01]
[2026-08-25 18.59.01] Status report: Tue Aug 25 18:59:01 IST 2026
[2026-08-25 18.59.01] PMS is stopped.
[2026-08-25 18.59.01] Databases are OK.
[2026-08-25 18.59.01] FTS indexes are OK.
[2026-08-25 18.59.01]
[2026-08-25 18.59.01] Starting PMS.
[2026-08-25 18.59.02] Started PMS
These would ALWAYS report corrupt after running a scan so looks like the server fix is good.
Hi, thanks for all the work done so far and for the future. I’m not entirely clear. Can we proceed with the latest version on GitHub? I see it’s been stuck for two weeks. Does that mean it’s ready?
However you can see this post from @OttoKerner to see what I’m talking about (orphaned accounts being retained in the database after being removed from Plex)
He knows how to do it. Perhaps you could collaborate about implementing a cleanup of this in your new program? Thanks for your work.
Edit: the option could say something like “prune orphaned users” and your program could cross check with Plex cloud which users are currently added, and prune the rest (orphaned). This should obviously not be included in the auto mode, but it should be safe in general.
The Powershell version for windows is still available to use on GitHub.
That is the Windows counterpart to the Bash script for the moment.
I’ve spent several full 8 hour days working on DBRepair (program)
over the past two weeks. It mostly works but there are some things I don’t like & a few edge-case faults which aren’t acceptable to me.
I spent all last weekend and all this week so far working on it.
It is almost ready to share again.
This is just my personal opinion of course but I think it should be a prune type command, but it should not be part of the auto sequence in case a user wishes to keep these orphaned records for some reason.
Thank you for all your hard work. I wasn’t sure the ps1 file was still working so I had been holding off for release of the application. Appreciate the continued efforts
I upgraded today to the latest version of the tool on github and got this when running sudo bash DBRepair.sh stop automatic status start exit
Database Repair Utility for Plex Media Server (Fedora Linux 44 (Workstation Edition))
Version v1.17.01
[2026-08-28 15.24.29] Stopping PMS. (60 second max delay)
[2026-08-28 15.24.30] Stopped PMS.
[2026-08-28 15.24.30] Automatic Check,Repair,Index started.
[2026-08-28 15.24.30]
[2026-08-28 15.24.30] Checking the PMS databases
[2026-08-28 15.24.40] Check complete. PMS main database is OK.
[2026-08-28 15.24.55] Check complete. PMS blobs database is OK.
[2026-08-28 15.24.55]
[2026-08-28 15.24.55] Exporting current databases using timestamp: 2026-08-28_15.24.30
[2026-08-28 15.24.55] Exporting Main DB
[2026-08-28 15.25.08] Exporting Blobs DB
[2026-08-28 15.25.13] Successfully exported the main and blobs databases.
[2026-08-28 15.25.13] Start importing into new databases.
[2026-08-28 15.25.13] Importing Main DB.
Error near line 5930821: malformed JSON
Error near line 5940946: malformed JSON
Error near line 5958360: malformed JSON
Error near line 5971630: malformed JSON
Error near line 6049803: malformed JSON
Error near line 6074956: malformed JSON
Error near line 6104496: malformed JSON
Error near line 6104619: malformed JSON
Error near line 6115970: malformed JSON
Error near line 6130675: malformed JSON
Error near line 6134673: malformed JSON
Error near line 6135060: malformed JSON
[2026-08-28 15.25.42] Error 1 from Plex SQLite while importing from './dbtmp/library.plexapp.sql-2026-08-28_15.24.30'
[2026-08-28 15.25.42] Cannot continue.
[2026-08-28 15.25.42] Repair failed. Automatic mode cannot continue. Please repair with individual commands
[2026-08-28 15.25.42]
[2026-08-28 15.25.42] Status report: Fri Aug 28 15:25:42 IST 2026
[2026-08-28 15.25.42] PMS is stopped.
[2026-08-28 15.25.42] Databases are not checked, Status unknown.
[2026-08-28 15.25.42] FTS indexes are not checked, Status unknown.
[2026-08-28 15.25.42]
[2026-08-28 15.25.42] Starting PMS.
[2026-08-28 15.25.42] Started PMS
Needless to say I restored my backup
EDIT: I should have said, I don’t think this is a problem with the tool, I think it might be a problem with pms itself as I exported those lines reporting malformed JSON and its all related to fts4 tables. Anyhow, I’ll probably start a dedicated thread on it but if anyone else using 1.43.4.10903 is willing to try and see if they get the same error, just make sure to have a backup