Server Version#: 1.32.5.7349-8f4248874
Player Version#: Plex Web 4.115.2
I have just done a lot of testing using wireshark and streaming from a remote Plex Server to a wifi-client.
Frequently I was seeing stream halting with RST packets getting sent by my client because too many packets were sent out of order , or never delivered from the server. Time and time again I checked connection settings, brought clients onto the local LAN, and fiddled with TCP settings.
I looked closely at the packets and saw that the DNS requests were being made to COX ISP DNS servers. These are of the format :
192-168-0-21.632902e1db4c40579439e305bc108501.plex.direct
so, they’re very easy to spot in a DNS query landscape. I noticed that whenever querying these I was having problems with the plex stream.
So, I swapped DNS servers on the router directly. Adding the following:
1.1.1.1
8.8.8.8
8.8.4.4
1.0.0.1
208.67.222.222
2001:4860:4860::8888
2001:4860:4860::8844
2606:4700:4700::1111
2606:4700:4700::1001
Once these Google, 1and1, OpenDNS servers were added , replacing the COX provided DNS servers I noticed and immediate change in the client performance. Also, reordered packets completely went away.
I can’t stress enough, change your DNS settings one the client side, this avoids the stream from being classified, and then having the stream mangled by COX.
I also want to reach out to PLEX DEVELOPERS to highlight this issue, because most people don’t know how to make these changes, or can’t make these changes. Please investigate an alternate means of providing the plex.direct name resolution – the 2 problems are :1 . the name has the IP in the name, so that’s not ideal, and 2. the direct response may be getting read by COX servers to mangle packets while the 2 IP addresses (source requesting the DNS resolution) and ( name resolved in the DNS call ).
Thank you. Now go use a different DNS provider everywhere !
