Can't upgrade past 1.40.2.8395

[media_plex_master]

Server Version#: 1.40.2.8395
Player Version#: 4.125.1
OS: Ubuntu 20.04.6 LTS

I haven’t been able to upgrade past version 1.40.2.8395, I have tried all the following

1.40.3.8555
1.40.4.8679
1.40.5.8854

I can revert to 1.40.2.8395 or any previous version and it works just fine. I have also ran ChuckPA PlexDBReapir. When trying to upgrade it gets stuck into, “Plex Media Server is currently running database migrations.” and I had times where the plex service would stop or be stuck in the running database migration. When tailing the log I see

Aug 15, 2024 09:25:34.237 [140280446126736] INFO - Running migrations. (EPG 0)
Aug 15, 2024 09:25:34.237 [140280446126736] INFO - Analyzing database.
Aug 15, 2024 09:25:34.368 [140280446126736] INFO - Vacuuming database.
Aug 15, 2024 09:25:35.847 [140280438799160] WARN - Crash: Crash reporting disabled
Aug 15, 2024 09:26:00.597 [140280446126736] ERROR - SQLITE3:0x80000001, 13, statement aborts at 7: [INSERT INTO vacuum_db.'blobs' SELECT*FROM"main".'blobs'] database or disk is full
Aug 15, 2024 09:26:01.442 [140280446126736] ERROR - SQLITE3:0x80000001, 13, statement aborts at 1: [VACUUM] database or disk is full
Aug 15, 2024 09:26:01.444 [140280446126736] ERROR - Exception thrown during migrations, aborting: sqlite3_statement_backend::loadOne: database or disk is full
Aug 15, 2024 09:26:01.444 [140280446126736] ERROR - Database corruption: sqlite3_statement_backend::loadOne: database or disk is full
Aug 15, 2024 09:26:01.444 [140280446126736] ERROR - Error: Unable to set up server: sqlite3_statement_backend::loadOne: database or disk is full (N4soci10soci_errorE)

Logs.zip (963.4 KB)

[ChuckPa]

@media_plex_master

Look at what it’s telling you.

The disk/partition where PMS is being stored – IS FULL.

The default location is /var/lib , which is the / filesystem

When you type df -h, it will show you.
You’ll see it’s full.

This is probably the time for you to move your PMS to a different partition or disk.

Aug 15, 2024 11:22:22.388 [139837734210192] ERROR - SQLITE3:0x80000001, 13, statement aborts at 7: [INSERT INTO vacuum_db.'blobs' SELECT*FROM"main".'blobs'] database or disk is full
Aug 15, 2024 11:22:23.250 [139837734210192] ERROR - SQLITE3:0x80000001, 13, statement aborts at 1: [VACUUM] database or disk is full
Aug 15, 2024 11:22:23.252 [139837734210192] ERROR - Exception thrown during migrations, aborting: sqlite3_statement_backend::loadOne: database or disk is full
Aug 15, 2024 11:22:23.252 [139837734210192] ERROR - Database corruption: sqlite3_statement_backend::loadOne: database or disk is full
Aug 15, 2024 11:22:23.252 [139837734210192] ERROR - Error: Unable to set up server: sqlite3_statement_backend::loadOne: database or disk is full (N4soci10soci_errorE)

I’ll help you do that. It’s very easy on Linux.

[media_plex_master]

I have plex on /dev/sdb1 (/plex) unless I’m missing something.

Filesystem                                                            Size  Used Avail Use% Mounted on
udev                                                                   35G     0   35G   0% /dev
tmpfs                                                                 6.9G  1.4M  6.9G   1% /run
/dev/mapper/ubuntu--vg-ubuntu--lv                                      20G   16G  2.7G  86% /
tmpfs                                                                  35G  4.0K   35G   1% /dev/shm
tmpfs                                                                 5.0M     0  5.0M   0% /run/lock
tmpfs                                                                  35G     0   35G   0% /sys/fs/cgroup
/dev/sda2                                                             974M  211M  696M  24% /boot
tmpfs                                                                  40G     0   40G   0% /media/ramdisk
/dev/sdb1                                                            1007G  459G  498G  48% /plex
/dev/loop2                                                             64M   64M     0 100% /snap/core20/2318
/dev/loop1                                                             56M   56M     0 100% /snap/core18/2829
/dev/loop0                                                             56M   56M     0 100% /snap/core18/2823
/dev/loop3                                                             92M   92M     0 100% /snap/lxd/24061
/dev/loop4                                                             92M   92M     0 100% /snap/lxd/29619
/dev/loop5                                                             64M   64M     0 100% /snap/core20/2264
/dev/loop6                                                             39M   39M     0 100% /snap/snapd/21465
/dev/loop7                                                             39M   39M     0 100% /snap/snapd/21759
tmpfs                                                                 6.9G     0  6.9G   0% /run/user/1000

under /var/lib

lrwxrwxrwx  1 root      root        21 Nov 23  2020 plexmediaserver -> /plex/plexmediaserver

Edit: Increased the /dev/mapper/ubuntu–vg-ubuntu–lv, which plex is now working, I ignored the disk is full msg because I didn’t think that would matter since plex lives under a 1TB mount /dev/sdb1 /plex.

[ChuckPa]

Don’t use symlinks.

Use the systemd override.
This means PMS is looking at the right places and not putting things into /var/lib to cause problems as is being reported.

/var/lib/plexmediaserver is Plex’s HOME directory.
If that shows full, it will get cranky

Also do check the free space available. Something is out of space.

Linux Tips

Customizing your Plex configuration on systemd based systems This information is intended for those systems where cat /proc/1/comm returns systemd. On systems which use Systemd (Ubuntu 16+, Fedora 26+, Centos 7+, Debian 8+,) you create this customization file which SystemD reads prior to launching PMS. This file is independent of the default configuration and service control mechanisms in /lib/systemd/system/plexmediaserver.service. Your customization override is preserved across updates, un…

[ChuckPa]

Supplemental:

This is my configuration

[chuck@lizum ~.2051]$ cat /etc/systemd/system/plexmediaserver.service.d/override.conf 
[Service]
Environment="PLEX_MEDIA_SERVER_APPLICATION_SUPPORT_DIR=/usb/plex"
User=chuck
Group=chuck
[chuck@lizum ~.2052]$ 

/var/lib/plexmediaserver still exists, is owned by my username, but is empty.

/usb/plex is really a 2TB SSD now. No longer a HDD

[media_plex_master]

Thanks, yeah i’ll just make a override.conf, I forgot about the location of /usr/lib/plexmediaserver. But either way seems resolved once I expanded /dev/mapper/ubuntu–vg-ubuntu–lv.

[ChuckPa]

/usr/lib/plexmediaserver is for the software. It is managed by the package manager.

/var/lib/plexmediaserver is where plex’s $HOME is.
We only need concern ourselves about this directory

[media_plex_master]

Yeah, what I just cant figure out why it was saying disk full when I had symlink to /dev/sdb1, for /usr/lib/plexmediaserver. I would of thought 2GB left would be plenty for /dev/mapper/ubuntu–vg-ubuntu–lv, since plex data isn’t technically running on main drive but the secondary drive with over 500GB of space. But yeah once i added additional space to /dev/mapper/ubuntu–vg-ubuntu–lv it worked instantly.

Either way I created an override.conf instead of a symlink.

[ChuckPa]

The symlink was the problem.

df will not follow the symlink to tell you the space free at the destination without using the trailing slash. PMS does not add trailing / to pathnames because it already knows it’s a DIRECTORY.