Bug: playback error after server upgrade

Server Version#: 1.24.5.5173
Player Version#: 5.0.1

After upgrading the server today the player always shows “Playback Error, An unexpected playback problem occurred.” even for videos previously played on the swme client but with a previous server version.

Tried all players I have and it works on iOS, but fails on LG, Chromecast (from iOS), Plex for Mac.

I’m using the latest Beta without issue

MacOS 11.6.1

LG Client 5.28.2

Upgraded to latest beta of the server, still having all the same issues. All players are upgraded, from what I can see.

Plex says tv client is 5.28.2, lg store says 5.0.1

That’s a loader only and is updated every now and then, to find your version of the Plex clients.

Server Setting/authorised devices.

Now what i would try is pick on a small library and renew the path on the server.

I can play all videos on my ipad and iphone plex players. I ran an optimize for tv to see whether it’s a codec issue and I get the same results - works on ios, fails on web, tv, chromecast. The latter is the most weird - the video starts playing on the phone, I select casting and on chromecast on tv it says “Sorry! Something went wrong”

Finally got around to debugging this. Logs say:

Dec 07, 2021 20:50:52.642 [0x70000726c000] ERROR - SQLITE3:(nil), 1, no such table: metadata_item_setting_markers in "select metadata_item_setting_markers.id as 'metadata_item_setting_markers_id', metadata_item_setting_markers.marker_type as 'metadata_item_setting_markers_marke
Dec 07, 2021 20:50:52.642 [0x70000726c000] ERROR - Soci Exception handled: sqlite3_statement_backend::prepare: no such table: metadata_item_setting_markers for SQL: select metadata_item_setting_markers.id as 'metadata_item_setting_markers_id', metadata_item_setting_markers.marker_type as 'metadata_item_setting_markers_marker_type', metadata_item_setting_markers.metadata_item_setting_id as 'metadata_item_setting_markers_metadata_item_setting_id', metadata_item_setting_markers.start_time_offset as 'metadata_item_setting_markers_start_time_offset', metadata_item_setting_markers.end_time_offset as 'metadata_item_setting_markers_end_time_offset', metadata_item_setting_markers.title as 'metadata_item_setting_markers_title', metadata_item_setting_markers.created_at as 'metadata_item_setting_markers_created_at', metadata_item_setting_markers.updated_at as 'metadata_item_setting_markers_updated_at', metadata_item_setting_markers.extra_data as 'metadata_item_setting_markers_extra_data' from metadata_item_setting_markers join metadata_item_settings on metadata_item_settings.id = metadata_item_setting_markers.metadata_item_setting_id where metadata_item_settings.guid = ? and metadata_item_settings.account_id = ? order by metadata_item_setting_markers.start_time_offset, metadata_item_setting_markers.end_time_offset

Created a small library, doesn’t help.

Now I can’t even delete it, as I get:

Dec 07, 2021 21:05:17.116 [0x7000071e9000] ERROR - SQLITE3:(nil), 1, statement aborts at 110: [delete from metadata_items where metadata_items.id in (select metadata_items.id from metadata_items  where metadata_items.id in (select metadata_relations.related_metadata_item_id fro
Dec 07, 2021 21:05:17.117 [0x7000071e9000] ERROR - Exception inside transaction (inside=1) (/data/jenkins/server/1787040707/Library/Scanner/LibrarySection.cpp:466): sqlite3_statement_backend::loadOne: SQL logic error
Dec 07, 2021 21:05:17.117 [0x7000071e9000] ERROR - Thread: Uncaught exception running async task which was spawned by thread 0x700006a3c000: sqlite3_statement_backend::loadOne: SQL logic error

From what I can gather, the metadata database is corrupted. No idea how to force it to be regenerated - I have updated the server today, which didn’t help.

The odd part is still that playback on iOS works (seeing the same errors), but on Chromecast and LG it fails.

Running “Optimise Database” fails with:

Dec 07, 2021 21:23:43.657 [0x7000070e3000] ERROR - [Database optimization/com.plexapp.plugins.library.db/DatabaseFixups] SQLITE3:(nil), 1, statement aborts at 81: [delete from tags where tags.tag_type != ? and tags.tag_type != ? and tags.id in (select tags.id from tags left join taggings on taggings.tag_id=tags.id group by tags.id having count(ta
Dec 07, 2021 21:23:43.657 [0x7000070e3000] ERROR - [Database optimization/com.plexapp.plugins.library.db/DatabaseFixups] Exception inside transaction (inside=1) (/data/jenkins/server/1787040707/Library/DatabaseFixups.cpp:216): sqlite3_statement_backend::loadOne: SQL logic error
Dec 07, 2021 21:23:43.657 [0x7000070e3000] ERROR - [Database optimization/com.plexapp.plugins.library.db/DatabaseFixups] Fixup 41DatabaseFixupsDeleteUnusedTagsAndTaggings threw an exception: sqlite3_statement_backend::loadOne: SQL logic error
Dec 07, 2021 21:23:44.385 [0x7000070e3000] WARN - [Database optimization/com.plexapp.plugins.library.db] Held transaction for too long (/data/jenkins/server/1787040707/Library/FullTextSearch.cpp:58): 0.670000 seconds
Dec 07, 2021 21:23:44.442 [0x7000070e3000] ERROR - [Database optimization/com.plexapp.plugins.library.db] SQLITE3:(nil), 1, statement aborts at 10: [DELETE FROM "main"."spellfix_metadata_titles_vocab"  WHERE id=684] SQL logic error
Dec 07, 2021 21:23:44.442 [0x7000070e3000] ERROR - [Database optimization/com.plexapp.plugins.library.db] SQLITE3:(nil), 1, statement aborts at 10: [delete from spellfix_metadata_titles] SQL logic error
Dec 07, 2021 21:23:44.442 [0x7000070e3000] ERROR - [Database optimization/com.plexapp.plugins.library.db] Exception inside transaction (inside=1) (/data/jenkins/server/1787040707/Library/FullTextSearch.cpp:71): sqlite3_statement_backend::loadOne: SQL logic error
Dec 07, 2021 21:23:44.447 [0x7000070e3000] ERROR - Thread: Uncaught exception running async task which was spawned by thread 0x70000628f000: sqlite3_statement_backend::loadOne: SQL logic error

Downloading the database produces a 22b “zip” file that doesn’t open.

It looks like I’m not alone: https://www.reddit.com/r/PleX/comments/qluzxn/error_occurred_loading_items_to_play/

I stopped the server, backed up the database and then ran the database repair procedure as outlined in Repair a Corrupted Database | Plex Support

And it works now.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.