For a couple of weeks my plex-server becomes ‘unreachable’ whenever I ‘Scan Library File’. However this unreachable status doesn’t impact anyone who is streaming when I get these occurrences. It is just seems to be the WC.
WC Version 4.8.4
PMS Version 1.18.2.2029
It Remains in this state for a minute or so before coming back.
Logs can be provided.
Thanks for reading.
I am having the same issue. Sometimes it requires a stop and start of the plex server service. It appears to affect all clients however as stated if a client is playing something then that continues to play fine. This appears to be happening from v1.18.0.1846 upwards however rolling back to 1.17.0.1841 does not fix the problem.
Current versions:
Server: 1.18.2.2029
Web: 4.8.4
Windows App: 1.4.0.924
I also have this problem with system freezing. I have narrowed it down to the plex media scanner that is gobbling up memory. This eventually causes the OOM killer to start killing off essential services.
I am running a plex-pass docker container on Ubuntu 19.10. I have a fairly large music library so I can recreate and demonstrate the problem really easily as follows: -
With my plex media server running, I open plex media player on windows 10. I then click on the activity icon(top right) and open the dashboard, scrolling down shows a static level of memory usage on the RAM graph. If I now click the home button followed by the three dots on my music library I can select ‘Scan library files’. I now return to the dashboard RAM graph and watch it, the memory used starts to rise and continues until the system freezes.
While I understand that the scan needs to use as much memory as it can, surely the server should cap what it uses to a level that does not cause system instability. I have succeeded in capping the memory available to the plex docker container and this prevents my system from freezing but the plex media scanner process gets killed by OOM-kill. The only way to fix this is to restart the container.
I think the new music scanner is the source of this problem, and a few others. It seems to be relentless in it’s search for matches, metadata and whatever else it does, sometimes appearing to repeat the process several times.
I call it a bug, but it seems Plex considers it a feature.
I am having this issue as well, in a big way. Literally the second a scan is initiated, the web gui goes down as unreachable. The server functions as normal otherwise though, accessible via clients to play and browse media, and even the new media is available with metadata, so it is still adding and tagging correctly during these “outages.”
I’m glad i’m not the only one. i have multiple servers, all on the same versions, but this only happens to one of the servers. The server goes down for anywhere between 20 seconds to a minute or 2. it’s driving me crazy. when the server reconnects, it scans everything just fine.
Ultimately I rolled back versions until I found one that worked, and it seems like the last iteration of 1.17 is the last time my server is stable. I rolled back to 1.17.0.1841-d42cfa161-ls55 and I guess I will just keep it there for a while.
any updates on this… I am having the same issue.
Unfortunately not. I’m still experiencing the issue, even after a series of PMS updates since my original post.
Same issue for me. 
I watched it and it is after the file scan when it tries to match … total lock out . items that are playing keep playing but new connections are not allowed. Going to start to see if it is one of the agents.
This was my only solution too.
This behavior appears on Mac PMS v1.18.x as well.
I’m rolling back to v1.17 on jeffp1717’s suggestion and hope to see this resolved in future releases.
I am looking into this in another forum thread here
I have only get one set of diagnostics and I would like more to look for commonalities
I am also trying to establish which version the problem started with?
https://support.plex.tv/articles/201643703-reporting-issues-with-plex-media-server/
https://support.plex.tv/articles/200250417-plex-media-server-log-files/
If you are unable to reach the server during the lockout, just copy out the Logs directory manually and zip the copy and upload here. Please do not zip the directory in-situ as that would skip the open files
Can confirm I too am having same issue since aroind the time of the last update.
Very frustrating.
While in web client I can the the server scanning and then it loses connection.
Sometime it comes back on it own but mostly I have to reboot it.
I have unchecked all scanning options for now to see if it makes a difference.
I moved my PC with plex direct to my modem and still does it. I do see CPU pinning at 100%.
I am also having this issue 
edit: going back to v16 seem to have fix the issue, ohh dear plex please fix this and get your house together.
I’m seeing it, too. Several weeks now. I’m worried it might be related to the size of my library. It’s ginormous, and running on an aging 2010 Mac Mini w/ 16GB RAM and High Sierra.
FYI - rolling back to 1.17 did the trick for me, too.
Thanks for reading.
