Server Version#: 4.69.1
Player Version#:all the ones I tried
I have done some extensive testing and opened problems with Verizon FIOS and also Firewalla support just to verify things like no double NATs. Neither seem to have any answers.
I have a plex server running on a windows 2019 server. For remote access, I have it configured to use port 32400 and not upnp. On my firewalla gold router, there is a port forwarding rule forwarding external port 32400 to my windows plex server. The firewalls on my windows/plex server are all disabled currently just to minimize possible causes. Once I enable remote access in plex, it appears to only stay available (green) for 1 minute and then goes red.
Just to verify the port forwarding rule was working, I created a very basic ftp server on a non standard port and duplicated the plex rule in my firewall with the exception of the port. I was able to connect to the ftp remotely so I know port forwarding is working correctly, just not for plex.
I opened problems with Verizon FIOS to see if I was somehow going through a double NAT and according to their 2nd level support, I am not. I also opened a problem with firewalla support and explained the whole setup and they don’t have any ideas.
Bottom line, it looks like the plex server opens port 32400 but only briefly and then never opens it again. So because of that, and what I was told by some other experienced plex users, this forces all remote connections to go through the plex relay, which unfortunately appears to force transcoding and therefore downsizing the video size when I know for a fact the bandwidth between my server and my remote users is more than enough to stream 1080p directly. They are using plex clients on devices (mostly firetv 4k sticks) which natively support 1080p and the x265 codec, and all the various audio codecs. No abnormal codecs at all.
There is nothing in between my plex server and the firewalla router. It is a 1 gig connection and my FIOS is also a 1 gig service.
Since I verified that port forwarding is working correctly with my ftp test, the only other thing I can think is that this is a plex issue.
No matter what I have tried (I also tried upnp) the remote connection is not staying available. I know this is a common issue and I have tried the workaround by setting a static port and created a port forwarding rule. My windows/plex server has a static ip address. This did not solve it.
Does remote access in plex have a timeout? Is it configurable? Is this a setting somewhere in an XML file or can the timeout which appears to be one minute, somehow be overridden?
I am really at a loss here and not sure what else to look at.
Oh. I also verified that all settings in the server and the clients are set to max or equivalent so they should be able to do 1080p with no issues.
That’s not how Plex works; when the server is running, it is always listening on TCP port 32400. Were that not the case, none of your local client would be able to connect. Of course, I’m making an assumption here: Your local client are able to connect, right?
Not in the sense that you’re suggesting. It’s more of a retry timer. It checks the status of remote connectivity availability every X amount of time (the specifics escape me right now). That’s likely why you’re seeing the behavior you are. When you first enable remote access it shows as successful by default. After the first real connectivity check it fails and the real status (not available outside your network) is shown.
Regarding the possibility of double-NAT, you can verify yourself whether or not you in such a configuration. First, browse to canyouseeme.org and note the IP address shown there. Then, in your Firewalla router, find the IP address which it shows for its WAN interface. Compare the two IP addresses; if they are not the same, chances are good that you are in a double-NAT configuration.
While you’re at canyouseeme.org, enter your Plex Media Server remote access port and have it check connectivity. This will show if an external tool is able to establish a connection to the port (and thus validate your port forwarding configuration).
Thanks for getting back to me. It does not appear I am beind a double NAT. The ip address on that WWW and in my firewalla match. One thing I did read in a plex article about plex remote is that if I have jumbo frames enabled, it may cause plex remote to fail. I currently do have jumbo frames enabled and will be changing that shortly and retrying. My remote connections are defaulting to using the plex relay and therefore limiting them to 2m connections and forcing transcoding, which I would definitely like to avoid. I will update here once I disable jumbo frames. Just can’t do it right now.
If you share server logs that may give some clues.
I work for VZ though not in Fios customer service and have the 1G service too. I do not believe we support Jumbo frames on the WAN for residential service. Jumbo frames on your LAN may work between devices in your home, but from doing speed tweeks on my network, I can say that anything over 1500 would fail a ping to the WWW for me. I would say you may have to disable jumbo frames on your router, and any TCP settings on your server and other devices should be left to the default frame size.
One other thing your might try, Enable remote access, save the settings then log the Plex server out and back in again and see if the setting holds and remote access stays available.
The plex server is virtual so I deleted the network adapter and created a new one just in case there were customizations that could break plex remote access. Remote access still won’t stay green. I also rebooted the switch it is connected to.
Are there any other options in plex that would break plex remote access? Like maximize/no limits options?
I am running out of things to try but I am bound an determined to get remote access working again. I will generate logs in a bit.
Which hypervisor do you use? Can you further clarify your environment; specifically as it relates to the virtual network configuration? Different hypervisors use differing terminology, but there’s almost always a distinction between NAT’d/host/bridged/etc… networking. That’s very relevant here.
It is an ESXi 7.x host and a windows 2019 virtual server. The network adapter is vmxnet3.
I tried disabling my anti virus just in case it actually monitors ports although there is nothing in the settings saying it does. And now remote access is green and seems to be staying green. There were some other things I tweaked as mentioned above like the nic, jumbo frames and rebooting the switch so at this point I am not sure exactly what is allowing it to go green. I am skeptical it will stay green though. Stay tuned!
So it seems to be better and staying green longer than before but it still seems to toggle back and forth.
Crystal ball says: “Information unclear. Problems are cloudy. Share logs.”
For funsies I’ve installed ESXi 7 (point something, update something) as a VMware Fusion VM on my Mac. And I’ve installed a Windows Server 2019 guest in that. It’s VMs all the way down.
It’s performing updates and such now. I’ll likely have a working environment to test in sometime today. Maybe. I’ve not used ESXi in a while and so wanted to be able to touch the virtual switch settings. The defaults have given the VM an address on my local network.
Once it’s all up-to-date, I’ll install a test Plex Media Server in it and do some remote access stuff.
Do not trust the Plex green dot. It can show you a moment of green when something downstream is blocking your port forwarding attempt. I know this because I was doing the same thing as you, when I discovered an error I had made which would have blocked every single packet. The moment of success you see could very well be a red herring as it was for me.
I would only test from outside your lan, with a tool like Open Port Check Tool - Test Port Forwarding on Your Router
Ugh. That makes trusting whether it is working or not very frustrating! Haven’t been home all day. Collecting logs now while it appears to go up and down and also while I enable/disable and click retry.
Logs are attached. Between 11:18EST and 11:45EST, remote access went green to red and I clicked on retry where it would stay green for a bit. I clicked retry several times and also enabled/disabled 2x toward the end. Hopefully the logs are helpful.
Plex Media Server Logs_2022-03-15_23-42-25.zip (5.9 MB)
If remote access is still working even when the green check mark goes from green to red and then back then there is no problem. The indicator is not reliable.
Why does this virtual machine have so many (3? More?) network interfaces and IP addresses? How many of them have gateways configured? (2?)
Mar 15, 2022 23:09:23.532 [5352] DEBUG - NetworkInterface: ignoring adapter 'Ethernet1' (10.90.90.2) because it has no gateway
Mar 15, 2022 23:09:23.540 [5352] DEBUG - Detected primary interface: 192.168.1.37
Mar 15, 2022 23:09:23.540 [5352] DEBUG - Network interfaces:
Mar 15, 2022 23:09:23.540 [5352] DEBUG - * 7 Ethernet {30EBA3DC-D18E-44AC-8B32-8283A408BF6B} (10.21.11.34) (loopback: 0)
Mar 15, 2022 23:09:23.540 [5352] DEBUG - * 12 Ethernet 2 {BE0E6618-6712-44D0-A53C-EEF0C5F934B5} (192.168.1.37) (loopback: 0)
The machine isn’t talking to itself on port :32400, which is … weird. Wrong. Broken. Plex should always be listening on :32400 locally, even if other ports are used externally. But it’s similar to something we’ve seen before - I have a small hunch.
Can you share screenshots of ALL the settings on System → Settings → Network?
I’m particularly curious about Custom server access URLs, but please share them all.
This is where it gets a little complicated. This machine also has a VPN client so that is the 10.21.* address. It also has a physical 10G cable connected directly to my Synology NAS with no network switch to provide a 10G connection between my “media” server and my NAS for doing backups. I just did a direct connection as it was the only 2 devices with 10G and it saved me some money from getting a 10G switch which are still a bit expensive, thus no router defined for the 10.90.* nic. . Otherwise my devices including my plex server would be backing up across a 1G link, much slower.
The plex server under network settings has the following network settings:
secure connections-preferred
No cert settings.
Strict TLS is disabled
Enable local network discovery (GDM)- enabled
Remote streams per user- 1
LAN networks - blank but have tried adding 192.168.1.*
terminate sessions - 60 seconds
Treat WAN IP As LAN Bandwidth - enabled
Enable relay- enabled
Custom server access URLs - blank
List of IP addresses and networks that are allowed without auth - blank
Webhooks - enabled
192.168.1.37 ethernet set as preferred.
Also just for clarity, plex is not using the VPN client. It is only there for downloading securely.
All of this was in place before switching to the firewalla router. I know it is not ideal. My plex server also happens to be my main download machine.
For whatever reason, remote seemed to stay available for quite awhile yesterday. Now it won’t go green at all. Assuming it being green actually means anything after reading some of the previous comments. So frustrating!
I have Fios…
The green light comes and goes, Its just an indicator, I ignore it…
If remote isn’t working I’ll get a text message…
