Individual libraries not loading since PMS update

Server Version#: 1.15.5.994
Player Version#: Web Client 3.99.2
OS#: Ubuntu 16.04.4 LTS (GNU/Linux 4.15.0-46-generic x86_64)

Plex Media (4.5 MB)

Hey,

I rarely have Plex issues but I seem to be hitting a few with the past few updates to PMS.

My last issue was detailed here.

That was resolved by downgrading from the Plex Pass version to the public release.

Yesterday, I added some content to my TV and Movie libraries. The update seemed to be infinite so I cancelled the scan to each library when I knew that all the new content had been added.

I upgraded my PMS (I think it was yesterday) to the latest Public release. Both my TV and Movie libraries then became inaccessible:

There was an unexpected error loading this library

I have several other libraries (which weren’t scanned for updates) and these remained visible.

So, I further upgraded my PMS to the Plex Pass Beta (detailed above) and the issue persisted.

I have restarted my Plex server to no avail.

I then scanned my database for corruption errors. The only error received was:

Error: no such index: index_title_sort_naturalsort

I ran a repair, and the corresponding error message was received:

Error: no such index: index_title_sort_naturalsort

I restarted Plex and checked. Both the libraries with errors were still not loading, whereas the others continue to be accessible. I started a ‘scan library files’ on the Movie library but I cancelled it as it was taking too long. Now, the Movie library is visible again, whereas, the TV continues with error loading.

I don’t have access to my ATV4K right now but it is running the latest TVOS and Plex app. Last night, it was unable to load the Dashboard. This occurred after the above problems.

One final thing, the Dashboard was visible at all times on PlexWeb and in it I could see the newly added content to the, then missing, Movie and, continued to be missing, TV libraries.

Thanking whomever in advance,

Antonio

Since the above, a new TV scan brought back my TV library. Although the scan took about 12 hours. It also added the new content it was waiting to do.

So, I thought I would be patient with a movie scan. 48 hours later it was still scanning and had only added a handful of the new content it was supposed to. I stopped it and started again but still no joy.

Here are the latest logs:
Plex Media Server Logs_2019-05-11_15-13-19.zip (3.4 MB)

The database is locked out most of the time

can you tell me if NFS is involved here ?
Or do you have the database open by another application?

May 11, 2019 07:02:24.435 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:25.444 [0x7f19ca6ce740] ERROR - Failed to begin transaction (../Library/Dir.cpp:40) (tries=7): Cannot begin transaction. database is locked
May 11, 2019 07:02:27.164 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:28.175 [0x7f19ca6ce740] ERROR - Failed to begin transaction (../Library/Dir.cpp:40) (tries=8): Cannot begin transaction. database is locked
May 11, 2019 07:02:29.628 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:30.644 [0x7f19ca6ce740] ERROR - Failed to begin transaction (../Library/Dir.cpp:40) (tries=9): Cannot begin transaction. database is locked
May 11, 2019 07:02:32.562 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:33.577 [0x7f19ca6ce740] ERROR - Failed to begin transaction (../Library/Dir.cpp:40) (tries=10): Cannot begin transaction. database is locked
May 11, 2019 07:02:33.581 [0x7f19ca6ce740] WARN - Caught exception while scanning Movies: Cannot begin transaction. database is locked
May 11, 2019 07:02:34.622 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:36.870 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:39.105 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:41.339 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:43.568 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:45.806 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:48.036 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:50.270 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:52.506 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:54.735 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:56.983 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:02:59.216 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:01.455 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:03.694 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:05.933 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:08.168 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:10.402 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:12.642 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:14.877 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:17.113 [0x7f19ca6ce740] WARN - Waited one whole second for a busy database.
May 11, 2019 07:03:18.326 [0x7f19ca6ce740] ERROR - Exception thrown while scanning: sqlite3_statement_backend::loadOne: database is locked

Thank you so much for taking time to assist me. The Plex Server is situated on a Bytesized server. Me not knowing how I would have the database open with another application should mean that I don’t (?)

I reinstalled the public PMS and it seems to be updating the stuck library now.

Here are the new logs:
Plex Media Server Logs_2019-05-12_18-04-42.zip (7.2 MB)

Could have it been a bad update installation or is this, coupled with the other issue I had, something to be concerned about?

I reached out to support on Bytesized. They had this to say:

No there is no NFS, everything is local attach storage, a simple harddisk :slight_smile: There is also no app we support uses a direct connection to the Plex database.

Usually when you see database already opened messages it means the app is leaking a connection that is never closed from within the application.

You mentioned you attempted a db integrity check/repair earlier - so you were opening the database with another program - was just wondering if it was on the live db as opposed to a copy of the file and if the session was left inadvertently for some time with the db file open by the sqlite3 tool you were using to do the check / repair

I will go back and look again at all the 1.15.5.994 logs you provided

Was the first launch of 1.15.5.944 at 04:20:30 am PMS time (which equates to 02:20:30 am GMT) ?

And what version of Plex Media Servr were you upgrading to 1.15.5.994 from ?

Thanks for making that query

We have specific steps that need to be done to overcome naturalsort related errors

See https://support.plex.tv/articles/201100678-repair-a-corrupt-database/

Were these errors even when following those steps?

Thank you again for your replies. I shall attempt to answer your queries the best that I can.

You mentioned you attempted a db integrity check/repair earlier - so you were opening the database with another program - was just wondering if it was on the live db as opposed to a copy of the file and if the session was left inadvertently for some time with the db file open by the sqlite3 tool you were using to do the check / repair

The way that I attempted a db check/repair was via the link you provide in your second post:

We have specific steps that need to be done to overcome naturalsort related errors
https://support.plex.tv/articles/201100678-repair-a-corrupt-database/
Were these errors even when following those steps?

Yes, these errors occurred when I followed those steps for Linux. I believed that was the way to do the db check/repair and is the only method I have attempted. I had contacted Bytesized prior to performing this to ensure that SQLite3 tools are installed, which they confirmed. The only program I used to perform these checks were via LXTerminal. Apologies if that is what you meant by an app. If that qualifies, then yes I did use one. However, I waited for either a message to appear or for a new command line to be available before I closed the window. I believe I did everything right but I am a command line novice.

Was the first launch of 1.15.5.944 at 04:20:30 am PMS time (which equates to 02:20:30 am GMT) ?

I am unable to confirm that. It could be but then again… I do tend to do stuff like that either late at night or early in the morning and 02:20 GMT would line up with that.

And what version of Plex Media Servr were you upgrading to 1.15.5.994 from ?

I had just upgraded to 1.15.4.994 but that itself didn’t seem to go swimmingly. Prior to that I had 1.15.3.876.

Latest status is the Movie library has updated though I had to eventually stop the scan as it continued above and beyond. But now, all posters in either TVOS, iOS, or PlexWeb need to reload and I am noticing a lot need to be added. The ones with missing artwork seem to correlate with movies that have had upgraded files in the past, of which I deleted the inferior version.

I think that answers everything(?) for now.

Thanking you.

Thanks for the replies

We do delete that index as part of the integrity check/repairs
"DROP index 'index_title_sort_naturalsort'"

I suppose if it did not exist, it should not matter for these steps and you should be able to proceed to the next step for the integrity check or repair

Did you stop at this point or proceed to the following steps?

I proceeded through to completion. The only other message I saw was when I was running the initial check with

sqlite3 com.plexapp.plugins.library.db "PRAGMA integrity_check"

And the message was “ok”

Otherwise everything came back with a command prompt, albeit some took longer for this to occur than others.

Thank you - so was ok with no corruptions detected

As a side note regarding database backups, I just noticed these in your logs

May 08, 2019 04:26:35.238 [0x7fc7c8ff9700] ERROR - Butler: Database backup path '~/.docker-conf/Library/Application Support/Plex Media Server/Backups' does not exist, skipping.
May 11, 2019 06:57:48.642 [0x7efebf7fe700] ERROR - Butler: Database backup path '~/.docker-conf/Library/Application Support/Plex Media Server/Backups' does not exist, skipping.

Did you specify this path for the backups of the database
~/.docker-conf/Library/Application Support/Plex Media Server/Backups
in the server scheduled tasks advanced settings ? or did Bytesized default to that? The default for backups is the same directory as the database - which would be
/config/Library/Application Support/Plex Media Server/Plug-in Support/Databases
if the default is changed, the specified directory must be created by the user to allow the backups to work

Yes, it carried on with no detected issues.

That was an issue on my part which has been rectified since I ran the database check/repair. Bytesized confirmed the correct location it should be a small few days ago and I amended it. Thanks for pointing that out though.

Thought I would say that my problems persisted so I decided to move my PMS to a local device and since I’ve done that, there have been no issues. So maybe it was my Bytesized server. Anyhow… thanks for all your assistance. Really appreciate you having taken time from your day to help me along the way.

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