I have such an odd networking issue with a weird very temporary fix. Searching for fix

Plex Media Server Remote Access
, ,
1

Server Version#:129.2.6364
Player Version#:

I’ll just be posting the reddit thread that I had posted since it covers a lot of the information and processes:

Edit 5: Well future peeps, get ready for the most flaccid update that you hate to see: it just started working for no apparent reaason*.* Yep, it’s the classic “it fixed itself” syndrome. I am deeply sorry on behalf of my server for every future troubleshooter. I’m sure this is like the fiftieth thread you’ve looked at and this one might even be the most similar one to your issue too! It’s honestly pathetic. I had seen easily over a hundred help threads and none of them helped. Now my very own thread that stood for justice of the future troubleshooters became something to through into the waste bin too. It’s poetic almost.

I guess to recap things I did learn and that might help your situation (it’s the least I can do after my “solution” after all were:

  • the error:
[Req#1594e] Missing X-Plex-Container-Size header. This will fail with status code 400 in the future.

was actually due to a custom NextDNS being on my phone and router. This didn’t fix my streaming at the time, but maybe it helped something else.

  • The “fix” in edit 2 for changing the port to 443 actually didn’t do anything. It seems like it was probably placebo. Currently, I’m back at the stock 32400 port
  • The “Broken Pip” issue in edit 3 still happens to this day, but it streams faster than I’ve ever seen it, so it’s either cosmetic, or doesn’t matter to my current situation.
  • You need to keep in mind that I wiped out my entire server with all the data and called “no taksies backsies”. My server was down for all the scanning and analyzing for days. The issue AFTER the wipe could genuinely have been it was too busy to send files to my phone. I DID have this issue before I wiped the server, so it was a real issue at a certain point.
  • Lastly, I know this is a complete coincidence (or is it? ( ͡` ͜ʖ ͡´)), but after installing jellyfin and trying to configure that disgusting monstrosity (I beg you don’t do it, I don’t care what the nerds say buying a domain name, setting up a reverse proxy and get that ■■■■ to work is not easy.) the server worked after I gave up on that project

So, maybe remote connection was the journey we made along the way.

Edit 4: More error codes that may offer help:

[HttpClient/HCl#8e] HTTP error requesting GET undefined (6, Couldn't resolve host name) (Could not resolve host: undefined)

[Req#1594e] Missing X-Plex-Container-Size header. This will fail with status code 400 in the future.

NAT: PMP, got an error: Not Supported by gateway.

Edit 3: A full factory reset of the router also did not fix the issue. Changing port to standard Plex port of 32400 also has changed nothing for performance of remote streams. The main error I’m facing is:

Failed to stream media, client probably disconnected after 2277276 bytes: 32 - Broken Pipe

If this error shows up, which in most cases it does, the stream is in buffer hell, or buffers and plays back a second or two before buffering for a while again on repeat.

Edit 2: I’ll edit this until I get a fix for everyone in the future. Everything works better under 443, but when I unplug and plug it registers the change in console, and then about a full minute or two later, it registers a second change with this:

NetworkInterface: received Netlink message len=92, type=RTM_DELADDR, flags=0x0
NetwrokInterface: Netlink address message family=2, index=2, flags=0x0
Network change.
NetworkInterface: Notified of network change (force=0)
Network change notification but nothing changed.

Edit: sub edit: Fixed some not permanent fix I might have fixed the issue. I changed to custom port forward I had to 443 instead and it seems to have fixed most of the problems. I guess time will tell if this was a proper solution. My external IP address had changed a bit ago, and I have been wondering if the ISP has had anything to do with it. Purely speculation from someone who knows hardly enough make assumptions.

Original Post:

My server has run great up until a month ago where remote usage had become spotty, and slow to load content, until this week where it’s for all intents and purposes not usable. I had no notable changes, but I decided it was time to do a clean install, since I was going to be upgrading the SSD anyways and didn’t need the watch history.

After THREE fresh installs of Ubuntu 22.04 LTS and THREE fresh installs of Plex and dealing with permission issues it seems that nothing has changed and I’m still having the exact same issue as before with remote playing working until I actually need to play something and I get stuck on buffering hell for seemingly forever.

I had my phone running on cell network for troubleshooting and it was stuck buffering while I was messing with the server, and I decided to change the LAN wire I had been using to my router to see if that was the issue.

As soon as I did that the phone started playing the video PERFECTLY, BUT this only lasted about 15 seconds and when I changed to the next episode it had the same issue. I tried the old LAN wire, and the next episode again played FLAWLESSLY for about 15 seconds which was enough time to buffer the whole episode.

This is endlessly repeatable and I cannot for the life of me figure out what is happening and why I’m facing this issue. The only thing the logs show is:

"Failed to stream media, client probably disconnected after 2277276 bytes: 32 - Broken Pipe"

This goes for ALL devices remote playing anything on different operating system. Does anyone have any idea what is going on?

Also to add: my networking has been consistently been 1000 down and 50 up through this period of time.

2

Some errrors when also trying to stream through plexamp show:

Failed to stream media, client probably disconnected after 229376 bytes: 32 -Broken Pipe

[Req#63d] Could not convert "state" ("error") to the correct type

[Req#701] Missing X-Plex-Container-Size header. This will fail with status code 400 in the future.
3

To be clear, the media does eventually play, but the playback time is extremely slow like if I have the worst connection in the world to the server, but my upload speed is solid at 50mps, my data pulls flac files fine down from other services, and it also worked well for an entire year until it slow turned into the sad state it is today.

4

Sounds like you might have some ISP introduced issues, however it will be good to revisit everything first just to confirm everything is setup and working properly. I would also advise switching back to 32400 for troubleshooting (or pick another non-standard port, using something like 443 may introduce it’s own issues since it is a common port used for other traffic).

Please read through Remote Access and Troubleshooting Remote Access support pages.
Then step through this guide: The Basics of Remote Access Troubleshooting

If you have been unable to get remote access working after you have completed all of the above (please be thorough and don’t skip any steps, or think you’ve already done all of this so you don’t need to bother), come back and we’ll dive into your setup.

5

Thank you for the help. I’ve looked through that guide thoroughly for the past few days, but just to say I did, I went through it step-by-step and everything looks good to me.

6

OK

So you’re running Ubuntu. Native install, Docker or…?
What’s your network setup? Eg. Are you using just your ISP provided hardware? Do you also have your own router / home network?
Any VPN?
Any firewall?
Is your Remote Access in PMS green and stays green?

7
  • I’m running Ubuntu on bare metal (Native)
  • I’m using my own hardware for the network. Coax>Modem>Router>LAN to Server with no additional bridges/routers in the home
  • No VPN
  • Firewall is enabled on the router, but was disabled for troubleshooting, same for the server. Disabled on both.
  • Remote Access always remains green and is always available from any device. The main issue is the endless buffering and the errors thrown up when it happens with “Broken Pipe” and “Connection reset by peer”. The device will buffer for a full minute and play 1 sec of video and repeat to the end of time. All devices have a robust amount of bandwidth to play with as well.

The content for testing has been a show with a bitrate of 900kb/s so it shouldn’t be an issue regardless of speed. Though, do note I have tried other media to ensure it wasn’t a faulty file.

8

Sorry, I was asking how your PMS is installed, but since you didn’t mention Docker I will assume native install for it as well.

OK, you have your own router as well as ISP modem. Just to double check and confirm that the modem is bridged and you don’t have any kind of double NAT situation? (Although given you’ve also said that your Remote Access actually seems to be stable, this should be moot, as would potential CGNAT issues.)

IPv4, IPv6 or both? If both, have you tried disabling IPv6 just to see if it makes a difference?

Are you using ISP DNS or custom DNS? If ISP, have you tried using custom public DNS like Google or Cloudflare?

Is it a single NIC on your server?

OK, interesting that your Remote Access seems to be stable… Is your Internet Upload speed set appropriately in PMS Remote Access settings?

What do your PMS Dashboard graphs look like when the issue is occurring? Is anything maxed out?

Are the items transcoding or direct play? (If possible, try and test both scenarios.)

Does this issue occur with local playback as well, or just remote? If just remote, does the dashboard show secure remote access during playback? Or something else eg. insecure, indirect, relay?

Sorry for the giant list of random questions. I’m just kind of brainstorming out loud at this point trying to think of potential things that could be related to what you’re experiencing.

9

Yeah, you assumed right. PMS is installed natively and is not containerized.

I may have phrased it incorrectly, but I own both the router and the modem for my network. I have IPv6 disabled on the router already and even tried with it on to see with no change.

I’ve also tried using the ISP DNS, Google, Cloudflare, and my own personal NextDNS, and nothing has changed unfortunately.

When looking at the graphs regardless of Direct Play, or transcoding, it show very little data transfer in the kilobytes when it’s stuck buffering. So it’s just not sending the data after the errors popup. The issue is purely Remote Play, and I have disabled relay to eliminate that possibility. All connections are made of a secure connection.

Thank you for typing all this out, I really appreciate this. Since I’m new, if we can’t figure this out, is there a way to get the attention of Plex staff, or is it just luck that I’ll get any support?

10

The only remaining item, that was buried in the middle, is: Is it a single NIC on your server?
The reason I ask that is it seems from my observations that anything beyond a single NIC where Plex is concerned can result in any manner of odd issues relating to connectivity and reliability. If it is not single NIC, then I would attempt that change to see if it helps. If it is single NIC, proceed below.

(Actually one other random thought: do you consider your router and modem powerful enough that they aren’t the bottleneck / potentially choking with the outbound throughput?)

And yes, the next step at this point seems to be to try and get some direct support. Often that can be a little luck but in this instance I would suggest attempting to summon ChuckPa into the thread. They are very knowledgeable especially on the *nix side of things. I don’t want to @ them until you’re ready though.

So, my suggestion would be to follow the steps outlined here https://forums.plex.tv/t/help-us-help-you-how-to-provide-logs-to-get-your-problems-identified-and-hopefully-resolved/672904 to generate a full log set, upload them here and then @ ChuckPa (alternatively if you don’t want to share your logs publicly, @ ChuckPa first and request to provide them privately). Hopefully they may have some existing knowledge with the errors you are encountering, or be able to discern something pertinent from your logs.

11

Sorry I honestly didn’t know what NIC stood for and forgot to look it up before replying. I know what the device is, I’ve just never seen the acronym for it before. I only have one built in to the motherboard option for the server. So that doesn’t seem to be the issue. I’ll reach out when I have more time to do everything they need me to do today, or tomorrow.

As far as the modem and router, they should easily be able to send the data and did so properly for a full year before this issue.

Seriously, though, thank you for your help with all of this. Your time is not going unappreciated.

1 Like
12

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