I love Plex but I'm considering removing it all together because I have to leave my PlexMediaServer service off pretty much at all times now. When Plex server is running it is flooding my network on ports 32412 and 32414 every second or so and causing anyhthing that I'm doing at the time to slow down. I'm not sure why it is doing this or how to fix it. Here is a wireshark pcap:
Just got home and looked at your wireshark capture file. It is not a problem, It is not even every second. It is every 5 seconds. That is normal, Nothing wrong with packets every 5 seconds. You only need to worry if they are thousands every second
It is a way for Plex Media Server to discover players and I believe it is related to the Enable GDM setting - but I never tried with that disabled as discovery of the server and players for remote / companion all use GDM as one of the means and some as the only means
See this support page which I got the Plex Team to update recently:
Whilst the support page title relates to firewall settings for ports, the contents of the support page is more than that
I'm still unsure why I'm having problems with my internet activities slowing down when PMS is running. Is there any way that you can help me narrow it down? Basically whenever I play online games that require a steady connection ever few seconds my connection to the game server completely slows down for about a second. It makes playing games while PMS is running pretty much unbearable. When I stop the PMS service on my server then everything is fine and nothing on my network is slowed down. Do you have any idea why this may be or are you able to help me narrow it down any further?
I'm still unsure why I'm having problems with my internet activities slowing down when PMS is running. Is there any way that you can help me narrow it down? Basically whenever I play online games that require a steady connection ever few seconds my connection to the game server completely slows down for about a second. It makes playing games while PMS is running pretty much unbearable. When I stop the PMS service on my server then everything is fine and nothing on my network is slowed down. Do you have any idea why this may be or are you able to help me narrow it down any further?
Maybe it is cpu and to do with activities on the system. Is Indexing option enabled in Advanced Library Settings for Plex Media Server : Generate Media Index Files During Scans ?
It is not just network activity that needs to be looked into, I think when the problem arises you need to look at the Performance figures/graphs at the time.
You will need a linux person to help you. On windows I would look at Perfmon and the Process cpu utilisation and Network utilisation
The top command will show you if anything is chewing up the CPU. Similarly free -m will show you any memory starvation issues (a low free value is normal, as long as the used swap value is low). The nload command (which you'll probably have to install) will give you a nice graph of your network load.
Edit: The vmstat command can give you a handy snapshot that includes average CPU load and memory usage.
A quick network diagram would help - is your server involved in any way in your Internet connectivity? Is it also your router?
The CentOS box that runs the PMS is a different computer than the one that's doing the gaming so usage wouldn't be an issue. My server also has plenty of CPU usage and memory free anyways.
And no the server doesn't have anything to do with my connection.
My gaming computer and my CentOS box are both connected to a switch. The switch is then connected to my router / modem.
Time then to run Plex up and run nload. If it is the cause of your network issues you should see significant traffic to and from your server (measured in Mb/s). If you do see that then NetHogs can tell you the cause.
Also, you are sure that it's a switch, not a hub? Check the output of ifconfig to see if any collisions are reported, or any other errors. You should see something like:
I know this may be pulling up an old thread, but I wanted to put some thoughts out.
I just went through a similar issue. Whenever my plex server was plugged into my network, all network activity would go down. It would even overload and crash my TV!
I did a wireshark capture and noticed the same udp packets mentioned. I started killing services and restarting to see what was sending them. Turns out to be plex.
I was getting 1000's a second. I had to block them on my router just to look through it.
Well finally, i got frustrated with looking at it and decided to look at some other things. Noticed my hwclock and date was wrong...
Turns out, my CMOS battery had died, and I restarted the server...hw lost its time, plex lost its time...nothing could figure out the correct time and plex just started spamming udp.
I am seeing exactly the same issue except it is broadcasting UPD to 1.0.0.0 twice every 5 seconds on two servers.
This is an external network so do not want to see this behaviour and it is filling my firewall logs.
Is there any resolution to this?
You can ignore multicast group broadcast messages (GDM) for the rest.
As for broadcasting on a Class A , publicly routable, network address, proper configuration of the port’s VLAN will stop such broadcasts where the host connects to the network.
Surely it should not be issuing this broadcast to a class A in the first place? Is this not a Plex issue?
This is coming from a wireless connected server inside my network. It isn’t getting out but filling my firewall logs. No idea how to do port level configuration as you suggest.
