I thought it dropped remote access because the people that use my server lost connection at that time. I could be wrong on the exact time though. I only posted the logs to hopefully help the developers get to the root of the issue faster.
I am on version 1.13.2.5142 & everything is running perfectly for days now. All family members are streaming without issue. So it was definitely the later versions of PMS that brought the problems.
I have updates turned off for the time being. Hopefully a solution will be found quickly.
I will try downgrading to that version as well, thanks for the tip. I’m about done with putting hours and hours into this issue. I tried scripting my way out of it, but most of the time the server ‘loses connection’ (to plex.tv, not in actuality) without actually setting PublishServerOnPlexOnlineKey=false - so there is no way to trigger a script to make it connect again. Could just have it retry every 5 mins, but that’s just ridiculous.
Reading all these people with the same problems I’m guessing this is in a Plex issue, rather than something local. Please get it fixed.
I downloaded wireshark but don’t know what to do from there.
Thank you for the feedback. Whilst i understand that downgrading is the solution for you, unfortunately this does not help finding any bugs that may exist - if we cannot reproduce the same issue you had then anu bug that may be there will remain unsolved
It is possible that there is a problem at the plex end and looking at your logs from the 20th December there is evidence of the Plex Media Server not getting events from the pubsub server (used for connectivity checks) - but we do need to investigate each such case. Lets see more examples
- You launch wireshark
- select the active network interfaces. They show up on the screen and the active ones you can see mini chart of data traffic - if not obvious - select all network interfaces
- start capture (there is capture start button)
- go through disabling and enabling remote access using default port mapping ie without manually specify port ticked
- if it fails, wait a couple of minutes, refresh the browser screen on the remote access settings page (F5 button) and if still showing failure, stop the wireshark capture
- get the server logs
- save the wireshark capture to default file type (pcap)
- note down the time when it is stopped
- zip the pcap file
- send me by private message
and upload corresponding logs zip here
Is there something I can do to help without breaking the hearts of my five users from constant drops in connection?
If it’s not a ball ache then I’m happy to help.
Sure, if you have specific logging you want I’m more than happy to oblige. Only the last logging you wanted from clients was a bit of a weird request imho. That or if you mean logging from the specific web client I use to check every morning and couple of hours manually (because yes, that’s what it amounted to now) then sure. I guess I can try and copy that when I see the same issue again.
So far I’ve moved my PMS from the docker install to a seperate KVM machine, to see if docker was the issue. I put another NIC into my server, and passthroughed that to that specific KVM, to see if the Linux Bridge was the issue. I even went as far as to replace my USG (router). All without changing anything. I also tried scripting the PublishServerOnPlexOnlineKey to turn off/on every x-minutes, but that seemed to make things worse rather than better. Is there some kind of limit on requests @ AWS site you guys have put in, or rather Amazon has put it where they temp-ban based on IP or something?
At this point I’m more than happy to try out anything. I will probably reverse proxy my current Plex instance, so I can give my friends a direct link to the server… that’s currently the only “solution” I can think of.
I just rebuilt my plex server today, new windows vm, same IP as the old one, ports didnt change, plex says 32400 is accessible from the outside, but no one can connect. I thought it was me, but ive tried everything and know the settings are correct. If this is plex issue, do they acknowledge that anywhere that i can read? Do they even know there is a problem? I found a copy of the version macrezz rolled back to but it wasnt from the plex site, do they keep those on their site somewhere? I normally have plex auto udate itself, does it keep those downloads anywhere so you can roll back if needed?
It wont let me install it without uninstalling the latest, will that lose my settings/library configuration if i do that? I dont want to setup all my libraries again if possible.
The strange thing is i can just now see random people connecting and playing movies, but i myself, on my local lan cannot and a friend that lives in another state cannot either. Very odd issue and im not sure how to fix.
Any ideas?
Edit: I moved my old plex appdata over to the new server, and put it in the same location on the new server then downloaded the latest version and did the install. I do not know what version i was running before, i assumed since auto update was on it would have been the latest. If i hadnt moved the files over and just copied them i could have fell back to my old server. ugh.
Not sure if it’s the same issue as mine… If you want to downgrade, it’s probably wise to treat it as a server migration. More info @ https://support.plex.tv/articles/201370363-move-an-install-to-another-system/
As for you on your local lan not being able to find your server - sounds like you haven’t advertised the internal IP to plex.tv. Settings => Network => LAN Network. Fill in your IP and you’ll probably be fine.
Thanks for the reply, it says that if you leave it blank it uses the servers subnet which should be fine right? Everything is on the same subnet.
I put my 10.1.1.0/24 in that section just to try and i still cannot access my media with any app on roku or iphone, but a web browser seems to work, (but that worked before) but that wont help me on my TV.
Any other ideas, or do you think its an issue with the latest release?
Also, any idea if plex keeps the auto update downloads or older versions on its site anymore? I cant seem to find them.
Thanks
we do need to catch diagnostics and logs when the remote access drops - so there will be some pain
In your case, you thought that opening the new dashboard bandwidth usage screen was a factor - so hopefully that will be easy to capture logs for = but please disable verbose logging and just have debug logging and do validate the availability of remote access by using a mobile device with wifi disabled or on remote wifi to confirm it is down - having established beforehand that it was working ok before
If we have an issue with the pubsub plex.tv servers connectivity testing events for remote access then that will show within the server logs whilst having the server remote access settings open
If we have a situation where the Plex Web on an android mobile device gets through remotely but the android plex app does not, then i would like to have logs from the android app as well as the server logs and indication as to when it happened
See https://support.plex.tv/articles/201573117-android-android-tv-fire-tv-logs/ for how to get the logs zip from mobile android device and email it to yourself
These are logs from just now. @ 9 am I re-enabled remote access and turned on debug logging. Just now I noticed Remote Access gave the exclamation mark again, so somewhere in that 2 hour window things turned sideways again. It was reachable via plex.tv while inside my local network, but not coming from outside (I currently have no advertised IPs btw). Tested via mobile device. Both the web app as the Android app showed the same behaviour.
Plex Media Server Logs_2019-01-12_10-54-24.zip (3.4 MB)
Notice that I haven’t actually re-enabled the remote settings by pressing the retry button when I downloaded these logs - it was still in the same error state.
Thank you for the logs. I do see a problem where events from the pubsub server assigned to your Plex Media Server are not coming in for the connectivity tests.
I cannot tell which pubsub server it was as that information is only logged at start up or when remote is disabled and re-enabled.
We can tell which pubsub server you are using now from the server Preferences.xml file which you can send me by private message or you could check yourself and let me know the value stored in field PubSubServer - it is possible that the server assigned before would be the same server
I do see that you are using Tautulli to repeatedly test connectivity every 1 minute. There has been a long debate on that and my recommendation to you here was to set it at 30 minutes. We do have timing bugs to do with connectivity testing which could lead to loss of remote access and the more you exercise that bit code the more chance you may encounter that timing bug.
There is no evidence that this timing bug was encountered as the logs only cover 1 hour and that is because you have verbose logging enabled. Verbose logging is only needed for specific investigations and should only be set if advised to do so. The default should be to run with debug enabled and verbose disabled
I am assuming for now that you do not have any IP Address filtering or geo filtering on IP addresses - as that would be one way for the events to be lost.
Going forward -
-
change the Tautulli remote access checks to be less frequest or do not use that option. I have stated before that we do have a bug in this area where the server may end up with it no longer being reached - so repeatedly going through code that potentially could break remote access is not advisable
-
disable verbose logging on server settings and enable debug logging
-
confirm that you have no IP filtering or geographic filtering
-
let me know what PubSubServer is used and capture logs at some point after a server restart (10 minutes in) to see the various timings for the Pubsub servers available (Plex Media Server selects the fastest connection one) - i am hoping what was in use before is the same as used now and i can get that pubsub server looked into
Yes, I noticed I still had remote access checking enabled @ Tautulli. Turned it off. Noticed Plex being shown down again at around 13, and then again at 14.
13:06 - verbose logging off, retry network pressed
13:07 - noticed the retry failed – lost connection to the server via plex.tv (internal network)
13:08 - reloaded plex.tv/web – working again
14:00 - checked plex.tv/web on phone - remote access exclamation mark, but is reachable (internal network
14:02 - checked on laptop, same thing
14:02 - checked on plex.tv/web via 4G on mobile - server is reachable (seems like it’s only reporting it’s not)
14:02 - checked plex app on mobile - working, tried starting something - worked without a hitch
14:03 - refreshed plex.tv/web on laptop – still working, but still reporting it’s offline
14:04 - pressed retry - seemed to be working again
14:05 - site shows as unable to connect to server (without refreshing)
14:06 - refreshed again, server shows as connected again
14:07 - downloaded logs below
Plex Media Server Logs_2019-01-12_14-07-02.zip (3.9 MB)
PubSubServer=“82.94.168.50” PubSubServerRegion=“ams” PublishServerOnPlexOnlineKey=“1”
EDIT 13/1: I added the plex.direct address to my alternative IP addresses. Now behaviour seems to be that the exclamation mark is back, but it is reachable. Let’s see if that holds up.
EDIT 14/1: Last night, it did seem to have published itself correctly (without any interference from me), but this morning the exclamation mark was back. So far my *.plex.direct seems to be working though, not for ‘indirect connections’ behind a proxy, but then - who uses those anyway.
Atlantasteven:
I downloaded wireshark but don’t know what to do from there.
-
You launch wireshar.
-
select the active network interfaces. They show up on the screen and the active ones you can see mini chart of data traffic - if not obvious - select all network interfaces
-
start capture (there is capture start button)
-
go through disabling and enabling remote access using default port mapping ie without manually specify port ticked
-
if it fails, wait a couple of minutes, refresh the browser screen on the remote access settings page (F5 button) and if still showing failure, stop the wireshark capture
-
get the server logs
-
save the wireshark capture to default file type (pcap)
-
note down the time when it is stopped
-
zip the pcap file
-
send me by private message
and upload corresponding logs zip here
Edited by sa2000 to remove capture zip
packet captures should be sent by private message as stated
I Have my logs And WS data for you available when ready, my server went belly up after win 10 update so i’m pretty sure because since I had to fix my audio drivers before they re-patched it again has something to do with it.
@sa2000 Thanks fr your help after looking at the logs myself I decided that I would test the router and firewall theory, So I booted up and old Win2008R2 server that ran my old plex server and this logged in fine with no issues and I went through all the settings like for like and couldnt find anything different, so at a loss I uninstalled and then ran an older version to see if it would make any difference and still had the same issue!
So I ended up backing up my server config (Windows2012r2) and then rebuilt the server with a Repair and upgrade first, still the same issue occurred yet all other internet traffic and network was not an issue
Even tried through a VPN to rule out my UK ISP blocking but still it wouldnt connect)
So Ive performed a fresh install on the Server with minimal roles and install plex again and now it running fine! Try to overwrite the app data files with the previous but there were a lot of files that had really long file names and Windows didnt like copying them so opted to roll back and start fresh
From the time the OS was back up and running I put Plex back on and then added the libraries back in (I named the server the same as the old one) it took maybe a few hours to run through and regresh all the meta data for the movies and music but now seem to be much faster after the re-install and the only thing missing metadata and artwork are a few old things that I can do manually so all in all the rebuild seems to be the quickest way to recover (if data is always kept separate from the OS store)
Im writing this off as an issue with my OS (maybe a Windows update corrupted something network related.
What is weird is that this issue just appeared and pretty sure noPlex Media Server.zip (18.6 KB)
updates happened for a good week or so.
I will see if I can zip the logs up as I did turn on the verbose logging as you suggested if you would like to review them for investigation to see if you can find out the cause but I’m sorted now however the nerd in me would like to know the solution Plex media Server log in zip was from 09/01/19 prior to my fresh install of OS let me know if there are others you need
Ive also added the same log file from my currently working server to see if you can pick out anything different?Plex Media Server 130119.zip (232.5 KB)
Cheers Guys
I’m having the similar issues but mine happened after the win10 update I’ve already done everything and checked pretty much all I could, nothing has changed but the update, ran a Wireshark for 5 min I have that data too. was waiting for a answer but didn’t get any, hoping someone has any ideas… even changed my port forwarding to a different one to change it up. Here are my logs, took forever to finally get it to DL was having issues getting it to complete. TIA
Plex Media Server Logs_2019-01-13_22-06-53.zip (6.8 MB)
