Connection went from Remote to Indirect to unavailable

Plex Media Server Remote Access
1

Server Version#: 1.16.3.1433-359b06978
Player Version#: Whatever is hosted on app.plex.tv

I’m writing this from work, so I’m on a remote connection. I was watching a video and it was streaming fine. When it ended, I clicked around to pick another video and started playback. Then I noticed the image quality was particularly blotchy and figured it must be transcoding. So I check the dashboard and, sure enough, transcoding from 1080p to SD. I cancel playback and go into the server settings and notice that it then said I was on an Indirect connection. Weird, I literally finished a Direct Play stream 60 seconds ago. So I click on the remote connection settings and it changes again to saying my server is not available outside my network. Weirder. I click on the Retry button and it now says my server is unavailable. And it’s been this way for the last 20-30 minutes.

I’m running PMS on an Intel Bean Canyon NUC, on Ubuntu 18.04.2 LTS. I can SSH into the machine, so I know that the hardware is running fine, and that the my DDNS is updating correctly to allow the SSH connection. In fact, I just ran all the security updates and they all installed without a hitch.

I updated the PMS server just this morning. And streamed 4 or 5 videos from it before the sequence I mentioned above. Any thoughts as to how my Plex server would jump across the three different statuses in the span of less than 2 minutes?

2

Updating this to add that my network consists of a Netgear Orbi and Netgear cable modem. The cable modem is in bridge mode, so no double NAT issues. I can remotely connect to my Synology NAS which is where my media is housed, so there’s another successful remote connection. The NAS is connected directly to the Orbi with one Cat 6 cable, and connected to a Netgear switch with another Cat 6 cable. Both connections on the Synology are link aggregated. The NUC is plugged into the Orbi directly. While SSHed into my NUC, I can navigate into my Synology, which tells me the mount is still valid and active.

With all this, I’m reasonably certain that it’s not a hardware problem. And since I can remotely access my NUC and NAS, I don’t think it’s an internal or external network problem. That and I was streaming fine just minutes before the connection went down.

3

I would revert back to a previous stable version of PMS. I am still on 1.15 for reasons like this, but I am running Win10 i5 3Ghz. All has been great.

4

It appears you have a multi-path configuration which isn’t best practice and will confuse things (indirect showing). You need a “Default Gateway” set in the Synology for at least one of the connections. (this is the likely contribiting to the problem).

I use this architecture which should work for you.

     +-------+
     | modem |
     +-------+
        |
     +-------+
     |router |
     +-------+
        |
     +--------+
     | trunk- |
     | capable|
     | switch |
     +--------+
        ||  |
        ||  +-------- Other single-leg devices (wifi, tv, etc)
     +------+
     | Syno |
     +------+

The singlular gateway for everyone is the router. Modem is invisible.
Switch does all the work

5

Would a change in architecture need to be made if I’m not actually running PMS on the Synology NAS? The machine I’m using for PMS is on a singular connection to the Orbi router. I’m not actually trunking two ports on my switch. The Synology NAS has one connection to the Orbi and one connection to my switch, aggregated for load balancing. It’s set up this way for internal connections to speed up file copies, so any network device only has a single hop to get to the NAS. It’s been running on the same setup for months without any remote connection issues.

My network looks more like this:

             +--------+
             | Modem  |
             +--------+
                  |
             +--------+  +--------+  +-----------------+
     +-------| Router |--| Switch |--| Rest of Network |
     |       +--------+  +--------+  +-----------------+
     |            |          |
+---------+  +--------+      |
| PMS NUC |  |  NAS   |------+
+---------+  +--------+

Also FYI, I tried rolling back to plexmediaserver_1.16.3.1402 but I still can’t connect.

ETA: I also set the Default Gateway setting on the NAS using the bonded connection. Still no dice.

6

Move the PMS NUC to the switch. Now the architecture is preserved.
“Switch does all the work” (because everything is on that LAN)

You lose performance going through the router.
Are you out of ports on the switch?

7

have you considered that perhaps your network admins have blocked your streaming?

8

I actually used the last open port on my switch for the bonded connection on the NAS. I’m happy to do some cable re-routing when I get home. What confounds me was that it was working fine earlier in the morning direct playing several videos (as it has for several months), then suddenly went to Indirect and unavailable.

9

I work for a cloud company so the network is wide open. But as soon as I saw that the server was unavailable, I double checked on my phone connected to Verizon. Then I checked on my other phone connected to AT&T. No connection on either device.

1 Like
10

what does your remote access say @ Plex > settings > remote access

I humbly doubt it is anything with your network architecture if it was working previously, more like either pms crashed, your router somehow started blocking the stream and/or lost your upnp port opening so plex can’t connect to your server.

if your ssh connection is still working, you could simply try to restart PMS;

  • systemctl stop plexmediaserver
  • sytemctl start plexmediaserver

(yes I know you can do restart too, but I like to see that stop actually stops the process before starting it)

11

Well, I’m not able to get to that settings page since I can’t connect, but the last thing it said was that my server was not available outside my network. Which was ironic given that I was accessing that screen from… well, outside my network. Then I tried clicking the “Retry” button next to the port forward option and, from that point on, I couldn’t get in.

I just tried manually stopping and restarting the PMS service, but still no change.

No one but me can log into the router to change any configurations. And I have remote admin disabled for it (Russian & Chinese hackers scared me off here). Of note, I didn’t have a manual port forward set up as I was relying on UPNP to set up the port allocation automatically (yeah, I realize this is another avenue for hackers, I pick and choose my battles based on laziness). So I’m starting to suspect you may be right, and that something went awry on the router affecting the automatically configured port fwd.

12

I concur with “If it was working” then the challenge now is to find out what changed on your local side.

13

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