Plex Media server crashing on windows 10

Server has been crashing a lot recently.

I extracted last few lines from log before crash:
Sep 09, 2022 14:37:59.015 [21996] DEBUG - Extras: Checked extra ‘The Divergent Series: Allegiant: LA Strada Per Il Futuro (Italian)’ and found matched: 1 Media matched: 1
Sep 09, 2022 14:37:59.016 [21996] DEBUG - Loading information from media bundle for part ID 64383 from “C:\Users\ameel\AppData\Local\Plex Media Server\Media\localhost\5\41fec3214516ce8a92985f4af565359e8c464c1.bundle\Contents\Plex Movie”
Sep 09, 2022 14:37:59.018 [21996] DEBUG - Loading information from media bundle for part ID 173058 from “C:\Users\ameel\AppData\Local\Plex Media Server\Media\localhost\5\c75e6005f33a7c6b62482b853e0ba70e597be4a.bundle\Contents\Plex Movie”
Sep 09, 2022 14:37:59.018 [21996] DEBUG - Loaded metadata for Allegiant (ID 7724) in 262ms
Sep 09, 2022 14:37:59.018 [21996] DEBUG - Doing expensive tags write for ‘Allegiant’ because something changed.
Sep 09, 2022 14:37:59.111 [21996] ERROR - SQLITE3:0x5bb30e4e, 267, statement aborts at 87: [update tags set parent_id=:tags_parent_id,metadata_item_id=:tags_metadata_item_id,tag=:tags_tag,key=:tags_key,tag_type=:tags_tag_type,tag_value=:tags_tag_value,user_thumb_url=:tags_use

Grateful for any help/support on this.

Please provide the entire log.

I already sent a private message with the logs

I sent another private message with the latest logs before crash.

Please send me the entire zip of the logs.

Sent!

hello. PMS remains unsuable.

I sent 2 batches of logs but have not received any response.

I appreciate you may be busy and working on various things, and would be grateful if you could provide an acknowledgement that the message and files have been received and a rough timeframe when you can provide an update.

thank you!

If I get a copy of the logs within the next 60 minutes I can take a look and get back to you shortly, otherwise Ill respond sometime tomorrow.

My comment above about sending the entire zip file of the logs was actually aimed for you. I can’t tell what’s going on with just 1 file.

Did you run “clean bundles” manually? If so, was there a reason?

I like my bundles clean. Never had a problem until the most recent build.

Crashes every time it runs. I couldn’t figure out why the server was crashing so I disabled the autorun of “clean bundles” and it stopped crashing.

Now if I manually run “clean bundles”… Crash instantly

See also this thread: PMS keeps crashing on Windows since last update August 31

@Explorer_200 Your logs indicate PMS is unable to access more memory that it needs.

Can you enable this advanced setting, https://support.plex.tv/articles/201105343-advanced-hidden-server-settings/, LogMemoryUse set that to 1, which should log more info on memory usage. Recreate the crash and get me a new PMS log.

I have sent you a message with the zip of the logs. PMS has not crashed on the latest run (I launched in the last half our or so).

I’m seeing disk i/o errors in your log. That typically means an issue with your drive where the Plex database is stored. It’s not a constant error, so it could be your drive is slowly starting to die and PMS is just running into this due to it accessing the database a lot.

Do you have another local hard drive installed? If so you can try moving your Plex AppData folder there and see if the crashes stop.

Thank you for the prompt analysis. Interesting. I can’t find any issues on my disks except for one (which does not run PMS). From the logs, can you tell which disk is giving i/o errors? I’ll try move PMS onto a new hdd in a couple of days.

On another note, some files are not matching correctly, and from console I can see the error:
[Req#1dc3] SQLITE3:0x4eedec8, 267, statement aborts at 60: [insert into tags (parent_id,metadata_item_id,tag,key,tag_type,tag_value,user_thumb_url,user_art_url,user_music_url,created_at,updated_at,extra_data) values (:tags_parent_id,:tags_metad
[Req#1dc3] Exception inside transaction (inside=1) (E:\jenkins\server\3582595338\Library\Tag.cpp:224): sqlite3_statement_backend::loadOne: database disk image is malformed
[Req#1dc3] Exception inside transaction (inside=1) (E:\jenkins\server\3582595338\Library\MetadataItem.cpp:7525): sqlite3_statement_backend::loadOne: database disk image is malformed
[Req#1dc3] Activity: Ended activity d9511943-9710-47d1-a88c-afb6634bca40.
Thread: Uncaught exception running async task which was spawned by thread 12576: sqlite3_statement_backend::loadOne: database disk image is malformed

there should be similar issues in the logs provided. Do you know how I can fix that?

Whenever you’re encountering messages about “database malformed” it is tiem to try a database repair, or you may lose your whole database.

https://support.plex.tv/articles/repair-a-corrupted-database/

Now this database damage may have been very well caused by either a dying hard drive or having almost no free space left.
Check for both before attempting a repair.

Thanks for the reply on this one. I have already attempted repair by following instructions here: https://support.plex.tv/articles/repair-a-corrupted-database/ but issue still persists. There is at least 200GB left on the HDD.

I will try again, but grateful for any other potential fixes.

Check the status of your hard drive. Use the chkdsk command line tool that’s included with Windows.

Using CrystalDiskInfo can’t hurt either, to read the S.M.A.R.T. register of your drive CrystalDiskInfo - Crystal Dew World [en]

Thanks for the tip. I akready use crystaldiskinfo and identified 1 disk with issues, but not related to PMS. Is there a way to check from PMS logs?

PMS logs will only contain informations about PMS.
A hard drive issue will only show indirectly in there, if at all.

in relation to this error: sqlite3_statement_backend::loadOne: database disk image is malformed

I moved PMS files onto a new hdd, and ran PRAGMA integrity_check using Plex SQLite. Response is “ok”. What else can I do?