Version 1.23.3.4707 breaks server

Server Version#: 1.23.3.4707
Player Version#: Latest Web players and App Players
OS: Debian 10 ( Buster ) on kernel 4.19.0-16-amd64

After updating to 1.23.3.4707 my streams wouldn’t buffer anymore.
On closer inspection my CPU was at 100% usage the whole time, transcoding a 720p stream ( it’s a Intel Xeon 1240L v5 with an Quadro K420 GPU, should be plenty powerful for a 720p stream ).

After restarting the ‘plexmediaserver’ service the CPU usage was at 0 again.
When opening one stream immediately 2 cores spiked to 100%, while not sending any data to the player.
Closing the stream would not stop the workload, so the 2 cores remained 100% used even when not playing anything on any device.
When opening a stream again, another 2 cores spiked to 100%, but there was still no buffering.
After making sure all streams were closed again, there were still 4 cores used 100% of the time.

Every opened stream seemed to peg another 2 cores at 100%, until I had a 100% load on all cores of my CPU, all while not a single frame of buffered data was received anywhere.
( Tested this on the android app and web browser ).

Solution:
Downgraded Plex server to version 1.22.3.4523 - now everything works as expected: when no stream is playing the server has barely any load, and buffered data is received in the clients.
Workloads for 720p streams spike to 70% for a few seconds then back down.

Side note:
I tried disabling Upnp in router settings ( as I found on another thread ), that didn’t make a difference.

Is this just me?

Without debug logs it would be tough to say, but it sounds like you had a back log of scanner jobs that started due to the fix in the upgrade.

Upgrades will sometimes kick off maintenance tasks that will consume all CPU.

Thank you for your reply @pl_5309, but wouldn’t that imply that my server should be using 100% on all CPUs without me having started a stream?
Or are these scanner jobs maybe triggered because I started a stream?

Something I forgot to mention: I upgraded to 1.23 just last night, since then until this morning my CPUs had all been running 100%.

I’ll upgrade again later today and collect logs!
Thank you again for your quick reply!