Plex Freezing, UI Unresponsive, way too often

General Discussions
,
1

Hello,

I am very frustrated right now, and I’d like some help. I’m a lifetime subscriber, so I hope that I can expect some technical support.

The issue is that Plex UI gets unreachable, apps cant load any data from it, however it keeps writing logs like nothing happened. It even shows attempts to access it, and it shows responses but nothing actually loads.

I was running Plex under Ubuntu and it happened like twice a day.
Now I run it under Proxmox -> Debian 9 VM -> Docker and it happens anywhere from 5 minutes to 1 hour. It’s impossible to use it like this.

It has plenty of RAM. It never really uses more than 1Gb when this happens. And there’s at least 6GB ram free both in VM and host.

It should not run into IO problems, I monitored disk usage from the host, and it’s not under load.

CPU is obviusly also at 0-10% when all this happens. And when it happens Plex’s CPU usage drops to 0-1% like it’s just idling.

Nothing in the logs jumps out at me as an error or anything.

What may be going wrong? Please help.

UPDATE:

I was using “Network - Host” in docker container for plex, changed it to Bridge and forwarded port 32400 TCP\UDP, waiting for it to hang again to see if it helps in any way. But that’s grasping straws

UPDATE 2:

No, 2 hours in and it failed when trying to play a video.

Logs

2

Try tagging you post with the appropriate server OS and client types and the relevant techs will see it.

3

Thanks, tagged, I’m not sure what clients to tag it with because it is a server side issue that affects all clients.

4

Update 3.

Did not receive any reply in a few hours. Since I was struggling with this issue for about a year now and then, and for the last week most of all since it got unbearable, decided to just go with a clean plex install in an LXC container instead of Docker. As I started to suspect that maybe database got too dirty or corrupted at some point, during some updates or maybe when I was transferring it from one install to another. I will have to recreate my carefully crafted playlists from scratch… But oh well…

First impression - seems as fast and snappy as it was when I first installed it. Added libraries and even during scan I seem to be able to play videos without hanging the whole thing up.

We’ll see how it goes from here…

5

Just to make sure I have this right?

Proxmox -> Debian VM -> Docker -> PMS?

If correct, Why are you doing all just? You are begging for all kinds of problems. There is no net gain by making something complicated just because you can.

6

Thanks for your reply. However I don’t see this as a major problem, for the following reasons:

  • It was also freezing with bare metal Ubuntu install WITHOUT even docker! It’s just that it happened only 1-2 times a day. Which in the long run is NOT a smooth experience. When I moved it into docker frequency of freezing increased to unbearable, but it’s not like it did not happen with Ubuntu install
  • Fresh install of Emby worked FINE with this installation, with no issues and very fast and responsive
  • Everything else in docker works fine with this installation method, including another HDD and database intensive application - Nextcloud, it was able to scan 100GB of data and over 300000 files in 7 minutes, about the same as with bare metal install. Well, it does not do transcoding, sure, but loads IO and CPU quite a lot during sync
  • I know this approach works for a lot of people, running docker inside a VM under proxmox, because a lot of people like the convenience of Docker and Docker Compose, but installing docker inside Proxmox host is… well, I prefer to keep host as clean as possible, otherwise there’s no point in using virtualization at all.
  • I was expecting some performance degradation, but not complete freezing

I believe that the main problem was not the way it was installed, it could’ve just contributed to it. I do believe that it was just something with the database. What made the issue worse is probably that I moved Plex Library folder from and SSD onto an HDD, which probably was the cause of it working slower and being less responsive, and maybe somehow contributed to increasing freezing frequency.

Note that first I tried moving it into LXC container and connecting it with the same Library folder as before. It did not solve the problem. It got a bit more responsive but it was still freezing and bugging out. Only when I went for a clean install - that’s what fixed the issue. By clean install I mean - creating a new database and server and setting it all up again from scratch, adding all libraries, changing settings, etc.

I did not make it complicated just because I can, quite the opposite. I made it comfortable and easy to manage. Complex but not complicated. Its a simple setup which allows me to backup docker host easily, and Docker is something that’s easy to manage and install.

7

The best way to produce the logs would be from https://app.plex.tv/web

Go to Settings > Troubleshooting > Download Logs This will create a ZIP file that you can upload to this thread for retrieval. RAR files have had a malware scare recently so I am hesitant to look inside yours.

The logging level should be Debug but NOT verbose, configured under Settings > General .

As someone that has used “all of the above” in running Plex and not encountered your issues it will be either your server hardware or how much you have deviated from a default install. IE no VPN or reverse proxy in with the Plex Docker. If your library is exceptionally large you may have run into limits that may or may not be able to be addressed.
How about your server specs, have you done a memtest to rule out defective ram?

8

Here’s a plain text file

Sorry for this assumption, but I have a feeling you did not read my posts. I’ll repeat.

I don’t currently have access to the broken server right now, as I reinstalled it since, as I mentioned in my post above.

I did not use VPN and did not use reverse proxy. The new install is exactly the same as the old one, except that I recreated it from a blank slate, IE I did not copy over old Library folder.

Assuming that there’s something wrong with my hardware also has no logic, because I already mentioned that I reinstalled Plex from scratch in an LXC container with a clean Library, let it rescan all files again, and it works perfectly fine at the moment, with all the same settings. Also Emby did not have such issues as well, so that rules out any hardware issues.

Running into limits of a large library, I don’t know what are the limits? I have about 4-5 TB of media: shows, movies, music, photos. I don’t thinks it’s THAT large.

One thing I now remember is that plex was almost constantly rescanning a few music files over and over trying to find either metadata or images for them, which are non existent (soundtracks of a few games, some so old that you cant really find metadata online for them in usual places where plex looks for them). Not sure if that contributed to this or not.

All this leads me to believe that the problem is somewhere inside the Library folder. Database or something. Broken or corrupted or overfilled or I don’t know what else. But that’s what fixed the issue entirely - recreating it with a blank Library folder and setting all up again. I just hope I won’t run into this issue again after a few years, I don’t like to lose my playlists and all the ratings I’ve set for songs, custom renaming, etc. ALl that I’ll have to recreate again. Also all watch progress is lost, obviously.

Just for the sake of completeness, hardware is i7 2600K with 16GBs of RAM, all of which was used as a gaming PC before, and now as a server for quite a while. Sure it’s old, but it’s working just fine, and it’s more than enough for 3-4 and even more Plex streams.

9 
http://127.0.0.1:37777/:/plugins/com.plexapp.system/messaging/function/X0J1bmRsZVNlcnZpY2U6QWxsU2VydmljZXM_/Y2VyZWFsMQoxCmxpc3QKMApyMAo_/Y2VyZWFsMQoxCmRpY3QKMApyMAo_
Mar 15, 2019 20:11:21.378 [0x7fcb47fff700] ERROR - Error issuing curl_easy_perform(handle): 35
Mar 15, 2019 20:11:21.378 [0x7fcb47fff700] WARN - HTTP error requesting GET https://10-10-0-4.24f8959c398945fda50b834ff78a0a69.plex.direct:32400 (0, No error) (OpenSSL SSL_connect: SSL_ERROR_SYSCALL in connection to 10-10-0-4.24f8959c398945fda50b834ff78a0a69.plex.direct:32400 )
Mar 15, 2019 20:11:21.379 [0x7fcb47fff700] DEBUG - HTTP requesting GET https://213-141-150-28.24f8959c398945fda50b834ff78a0a69.plex.direct:21490
Mar 15, 2019 20:11:21.380 [0x7fcb6a7fc700] ERROR - Error issuing curl_easy_perform(handle): 35
Mar 15, 2019 20:11:21.380 [0x7fcb6a7fc700] WARN - HTTP error requesting GET https://192-168-1-19.24f8959c398945fda50b834ff78a0a69.plex.direct:32400 (0, No error) (OpenSSL SSL_connect: SSL_ERROR_SYSCALL in connection to 192-168-1-19.24f8959c398945fda50b834ff78a0a69.plex.direct:32400 )
Mar 15, 2019 20:11:21.380 [0x7fcb477fe700] ERROR - Error issuing curl_easy_perform(handle): 35
Mar 15, 2019 20:11:21.380 [0x7fcb477fe700] WARN - HTTP error requesting GET https://172-17-0-1.24f8959c398945fda50b834ff78a0a69.plex.direct:32400 (0, No error) (OpenSSL SSL_connect: SSL_ERROR_SYSCALL in connection to 172-17-0-1.24f8959c398945fda50b834ff78a0a69.plex.direct:32400 )
Mar 15, 2019 20:11:21.383 [0x7fcb47fff700] ERROR - Error issuing curl_easy_perform(handle): 7
Mar 15, 2019 20:11:21.383 [0x7fcb47fff700] WARN - HTTP error requesting GET https://213-141-150-28.24f8959c398945fda50b834ff78a0a69.plex.direct:21490 (0, No error) (Failed to connect to 213-141-150-28.24f8959c398945fda50b834ff78a0a69.plex.direct port 21490: Connection refused)

Running a custom Cert?

As for the processor utilization, having done hardware for a long time, the fewer layers of abstraction in the chain between application and actual hardware the faster it will run and more responsive it will be. I submit: Why waste CPU cycles when not absolutely necessary knowing the processor is already approaching vintage status? I run an i7-3740qm and can feel the difference immediately when using VM or container.

10

Supplemental.

Running on the bare hardware, or manually connecting all the linkages and udev privileges, would allow HW transcoding

Mar 15, 2019 20:11:22.824 [0x7fcb467fc700] DEBUG - Codecs: testing h264_vaapi (encoder)
Mar 15, 2019 20:11:22.824 [0x7fcb467fc700] DEBUG - Codecs: hardware transcoding: testing API vaapi
Mar 15, 2019 20:11:22.825 [0x7fcb467fc700] ERROR - [FFMPEG] - No VA display found for device: /dev/dri/renderD128.
Mar 15, 2019 20:11:22.825 [0x7fcb467fc700] DEBUG - Codecs: hardware transcoding: opening hw device failed - probably not supported by this system, error: Invalid argument
Mar 15, 2019 20:11:22.825 [0x7fcb467fc700] DEBUG - Codecs: testing h264_nvenc (encoder)
Mar 15, 2019 20:11:22.826 [0x7fcb467fc700] ERROR - [FFMPEG] - Cannot load libcuda.so.1
Mar 15, 2019 20:11:22.826 [0x7fcb467fc700] WARN - avcodec_open2 returned -1313558101 for encoder 'h264_nvenc'
Mar 15, 2019 20:11:22.826 [0x7fcb467fc700] DEBUG - Codecs: hardware transcoding: testing API vaapi
Mar 15, 2019 20:11:22.826 [0x7fcb467fc700] ERROR - [FFMPEG] - No VA display found for device: /dev/dri/ren
11

No SSL anywhere, no domains, nothing. Just LAN ip and plex’s own proxy stuff. Port forwarded through UPnP and then changed to manual forwarding.

I understand it, but it does not justify plex just becoming infinitely unresponsive. Slower yes, unresponsive no. Also - I tell this for the 3rd or 4th time - it was happening on bare metal Ubuntu install, just not as often, but also every day! Not from the start, but it started happening after 1-2 years of use. Then it got worse under VM+Docker. Now with clean install under LXC it works without any hiccups, a few days and not once did it freeze. I don’t see how it can not be evidence that problem was somewhere in it’s database or library folder.

Thanks for the vintage status :smiley: But to be honest I used the same CPU in my gaming rig up till just about 6 months ago, it was holding up quite well. Right now I see no difference between running Plex on bare metal Ubuntu vs an LXC container under Proxmox. There was also not THAT much difference in performance under VM and Docker, but of course there was. Also it rarely even reaches 100% load.

I may upgrade it to 3930K when I finally decide to retire my current workstation.

It’s i7 2600K, with no video chip inside, and there’s no GPU in the system. I don’t think it supports any of that. At least I dont even have /dev/dri on the host.

Why would I even need a GPU if CPU handles enough streams for me and family as is?

12

Been about 4 days now since I made a clean install of Plex, wiping the whole Library and database. Happy to report that this solved the issue, not once did it freeze in these days.

So I believe this is enough proof that problem was not in the way I was running it and neither in my ‘vintage’ CPU like you tried to imply, but in the database. Oh, by the way, a lot of people run Plex on i5 CPUs of 3-5 gen, i3, old amds and even RPi with success. i7 2600K can still easily outperform those, so I’m not sure where this even came from.

The only problem I have now is that I had to ‘remove’ my old server name, and created new server with the same name which apparently makes some clients confused. Had to clear browser’s cache twice before it stopped telling me that I don’t have access to the server (logout - login cycle always fixed it). Same with mobile apps.

So I suppose this can be closed. To anyone else who encounters the same issue, try a fresh Plex install and see if it helps.

13

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