Since PMS 1.29.2, there have been internal changes.
That’s what it’s tripping on – badly.
This isn’t a PMS error. It is an incompatibility with PMM.
Engineering has made a change to accommodate this in 1.32.8 (again, fingers crossed)
1 Like
@chazlarson @KSPL3X I can at least confirm that in the upcoming release (1.32.8) the 500 error is resolved. I don’t know if this is going to be the silver bullet for the overall issue being reported in this thread but the 500 error is resolved.
2 Likes
Since pmm is using Python-PlexAPI to make the curl Endpoint calls, is this hammering issue pmm or is it Python-PlexAPI? Just trying to understand where the fix should ultimately be applied to. The Python-PlexAPI that is used in pmm is here https://github.com/pkkid/python-plexapi
If you scroll back,
I don’t have time to go back through this entire post, but has PMM implemented the updated batch API which is now available in PlexAPI third-party library?
I would like to know if that helps or if the issue persists.
Not yet. Dev for PMM has committed to the update, it has not yet done so.
As bull has stated, the Dev has committed to making the change but as it is a single-dev project, it has yet to be implemented.
Whilst it may offer performance improvements, I would hate to put all the eggs in that basket when we still have the side of the issue where the same version of PMM (and in turn, Python-PlexAPI) used on two different PMS versions can show a significant slowdown in newer versions of PMS - this has been repeatedly tested and observed when using static versions of PMM.
A two-pronged approach to tackle both of these things can hopefully means PMM can see the same (if not better) performance than we originally saw on pre-1.3X versions
@ChuckPa any update on how the Engineers got on testing against PMM?
I have no feedback on that yet.
We’re in vastly different timezones and not yet caught up with each other.
@Atomatth I see that 1.32.7.7621 has gone live for all users, I assume that this does not include the 500 error fix?
Correct. 1.32.8 will have this fix and will be coming soon 
3 Likes
Thank you so much!
Aw, RIP OG Blizzard
1 Like
I have run this on the 1.32.8 beta and can confirm that the 500 error no longer appears.
Thanks again to all for the quick turnaround on this one.
We’re now back to the general performance issues raised in this thread.
1 Like
Running the same version of PMM, nightly22, from the same machine against different versions of Plex, here are my test results. PMS was running on the same server (unraid with hotio images) all within their own directories and all with the same library and setup. so independent DBs and folder structures. Same PMM run commands. Same PMM config to exercise Plex. Tests were not run in parallel but instead, when one finished, I shutdown PMS container, started the next one, and ran the same PMM command but now it’s pointing to the new PMS container running.
To give you an idea of the size of the library tested against has 8 movies and 5 shows with a total of 140 episodes.
All libraries had credit and intro detection off. Video thumbnail creation was also turned off. No one was watching anything on Plex during these tests. No additional shows or movies were added during these tests and Plex maintenance was not running.
Test results:
1.32.8 --> Run Time: 0:45:34
1.32.6 --> Run Time: 0:37:18
1.29.2 --> Run Time: 0:36:37
Seems like we still have a performance degradation. 1.32.8 is 9 minutes slower than 1.29.2 .
6 Likes
My main concern is that it hangs the plex database and produces the SQL Sleeping 200ms error. I don’t care if it’s a little slower as long as it doesn’t freeze up Plex anymore.
If I use PMM on any of the 1.30.x database hangs. When I revert to older version it proceeds fine.
1 Like
Please open up a #pmm-help on our Discord for more assistance. The Busy DB Sleeping 200ms is something that can be caused under several different situations. Some are not even related to PMM.
1 Like
Hey @ChuckPa just following up on this. We’ve noticed that the latest 1.32.8 beta fixes the 500 error, so thanks a lot for sorting that out!
We’ve now got a PR ready for PMM so that it utilises batch changes (pretty much) wherever possible, which is giving us good speed improvements.
Recently however, I’ve noticed some pretty big slowdowns on my server. It’s taking over a minute to add or remove a TV Show from a collection on v1.32.8, but if I rollback to 1.31.2 it’s almost instant - and I’m doing this via the Plex web app, not PMM.
Has any further investigation into the recent slowness been conducted?
3 Likes
Curious about the status on this as well.
1 Like
it would also be good to know when 1.32.8 is planning to be released on the Public branch, the beta has been out for 27 days so far… can’t wait to see the back of this issue.
