Ubuntu Server- PMS is started, but no web UI

I have PMS running on Ubuntu 16.04.5 LTS. PMS appears to be running, but is not responding when I go to [IP]:32400/web. It was running fine up until a few days ago. I installed using apt, and do automatic updates. Perhaps a recent update broke something? Please help.

bryan@plex:~$ sudo service plexmediaserver status
[sudo] password for bryan:
● plexmediaserver.service - Plex Media Server for Linux
   Loaded: loaded (/lib/systemd/system/plexmediaserver.service; enabled; vendor preset: enabled)
  Drop-In: /etc/systemd/system/plexmediaserver.service.d
           └─override.conf
   Active: active (running) since Thu 2018-08-23 16:37:42 MDT; 23h ago
  Process: 1357 ExecStartPre=/bin/sh -c /usr/bin/test -d "${PLEX_MEDIA_SERVER_APPLICATION_SUPPORT_DIR}" || /bin/mkdir -p "${PLEX_MEDIA_SERVER_APPLICATION_SUPPORT_DIR}" (code=exited, status=0/SUCCESS)
 Main PID: 1374 (sh)
    Tasks: 5
   Memory: 46.9M
      CPU: 398ms
   CGroup: /system.slice/plexmediaserver.service
           ├─1374 /bin/sh -c LD_LIBRARY_PATH=/usr/lib/plexmediaserver "/usr/lib/plexmediaserver/Plex Media Server"
           └─1377 /usr/lib/plexmediaserver/Plex Media Server

Aug 23 16:37:42 plex systemd[1]: Starting Plex Media Server for Linux...
Aug 23 16:37:42 plex systemd[1]: Started Plex Media Server for Linux.
bryan@plex:~$

1 Like

Here is my ovveride.conf:

bryan@plex:/etc/systemd/system/plexmediaserver.service.d$ more override.conf
[Service]
Environment="PLEX_MEDIA_SERVER_APPLICATION_SUPPORT_DIR=/media/storage/plexlib/Library/Application Support"

My library is on a separate, large drive mounted in /media/storage/ under the directory plexlib.

Move the library out of the /media directory and you’ll be a LOT better off.

Mount it in some place you create like /storage. /media is owned by Gnome, Nautilus, and the automounter.

Also, if it’s a physical large, locally mounted drive (mounted in /etc/fstab), you’re ok. If it’s a network mount, PMS will not work over the network (file locking requirements)

Hmm. This is a headless box so I shouldn’t need to worry about gnome stuff, right? I will move it though to see if that resolves it. It is a “local” drive. This is actually a VMware esxi guest.

-BW

Headless means nothing. ESXi means nothing.

Did you load desktop or server?

Do you have log files I can see? A tar.gz or ZIP of the Logs directory is fine. I can work with that

Ubuntu server. I’ll upload logs later.

-BW

Please restart the server and let it run about 2 minutes then stop it before you grab the logs.
If it sits idle too long, all the important information will have elapsed from the log buffer

I have the logs. Can I PM them to you instead of posting publicly?

Just sent you logs via PM. Let me know if I should post any portion here.

Bry,
It’s running and streaming to a ChromeCast. It is seeing the “WAN” designation indicating VPN. If not, the some modem/router is enforcing DNS rebinding protection against *.plex.direct (private) domain.

This ‘relay’ should not be occurring.

Aug 16, 2018 03:01:30.170 [0x7fa7f5bff700] INFO - [PlexRelay] Allocated port 25258 for remote forward to localhost:32401

There are additional steps to check however please verify DNS rebinding is not blocking normal operation and detection.

If you can establish a normal SSH tunnel with http port forwarding and see the Web UI, then the issue is network and not server.

To verify, without using VPN

ssh -L 8888:127.0.0.1:32400 -l server_username_here   ip.addr.of.server

Then open http://127.0.0.1:8888/web in a fresh in cognito browser tab.

Hmm, still no luck even with port forwarding over SSH. I think the last time it had successfully connected to my chromecast was 8/16, right? I haven’t been able to successfully stream anything since then.

I run DNSMASQ on my router (Ubiquiti USG). I haven’t explicitly blocked DNS rebinding, but I added rebind-domain-ok=/plex.direct/ to my dnsmasq config.

Rebooted router, etc. after DNSMASQ change. Rebooted Plex ubuntu server. Tried SSH port forward again, as well as from my desktop.

Right. 8/16 was the last.

I just resolved a case where there were Plex.tv auth problems because of UK ISP MTU limitations.
The default 1500 MTU was too big for the ISP. I’m reviewing your logs now to look for the same misleading “Jumbo Frames” error message

Posting a photo of my putty tunnel config, just to be sure we’re on the same page.

Accessing http://127.0.0.1:8888 does not work.

I also tried telnet-ing to localhost:32400 to see if anything was listening on the plex box, no response there either:

bryan@plex:~$ telnet localhost 32400
Trying ::1...
Trying 127.0.0.1...
telnet: Unable to connect to remote host: Connection refused
bryan@plex:~$

you SSH to the target host IP

Let’s step back.

  1. Ubuntu is headless. Is SSHD enabled and started?
  2. From putty, you SSH to the IP of the server

A few more thoughts. It seems that plex isn’t listening on 32400 at all for some reason.

bryan@plex:~$ sudo netstat -plnt | grep ':32400'

and

bryan@plex:~$ sudo lsof -i :32400

both return no results, yet the PMS service seems to be normal:

bryan@plex:~$ sudo service plexmediaserver status
● plexmediaserver.service - Plex Media Server for Linux
   Loaded: loaded (/lib/systemd/system/plexmediaserver.service; enabled; vendor preset: enabled)
  Drop-In: /etc/systemd/system/plexmediaserver.service.d
           └─override.conf
   Active: active (running) since Sat 2018-08-25 16:56:16 MDT; 56min ago
  Process: 1365 ExecStartPre=/bin/sh -c /usr/bin/test -d "${PLEX_MEDIA_SERVER_APPLICATION_SUPPORT_DIR}" || /bin/mkdir -p "${PLEX_MEDIA_SERVER_APPLICATION_SUPPORT_DIR}" (code=exited, status=0/SUCCESS)
 Main PID: 1384 (sh)
    Tasks: 5
   Memory: 46.9M
      CPU: 318ms
   CGroup: /system.slice/plexmediaserver.service
           ├─1384 /bin/sh -c LD_LIBRARY_PATH=/usr/lib/plexmediaserver "/usr/lib/plexmediaserver/Plex Media Server"
           └─1387 /usr/lib/plexmediaserver/Plex Media Server

Aug 25 16:56:15 plex systemd[1]: Starting Plex Media Server for Linux...
Aug 25 16:56:16 plex systemd[1]: Started Plex Media Server for Linux.

Yep, I’m using putty on a windows box to ssh into my headless ubuntu box. sshd is running fine, since I am SSH’d into it successfully. For good measure:

bryan@plex:~$ sudo lsof -i :22
[sudo] password for bryan:
COMMAND  PID  USER   FD   TYPE DEVICE SIZE/OFF NODE NAME
sshd    1360  root    3u  IPv4  16168      0t0  TCP *:ssh (LISTEN)
sshd    1360  root    4u  IPv6  16170      0t0  TCP *:ssh (LISTEN)
sshd    1629  root    3u  IPv4  24601      0t0  TCP ubuntu.brywhi.com:ssh->BWLaptop.brywhi.com:62570 (ESTABLISHED)
sshd    1727 bryan    3u  IPv4  24601      0t0  TCP ubuntu.brywhi.com:ssh->BWLaptop.brywhi.com:62570 (ESTABLISHED)

Do you have IPv6 enabled on the Ubuntu box? Looks like you do and it’s active

There is a known issue where PMS doesn’t setup correctly on 127.0.0.1 if IPv6 is enabled.

FYI for future reference- Chuck was able to help me get this fixed. There must’ve been something strange in my Preferences.xml. file. We nuked the current one, which let PMS create a new one after restarting the service, and it came right up. Here are the steps to resolve:

  1. Kill PMS service:
bryan@plex:/var/lib/plexmediaserver/Library/Application Support/Plex Media Server$ sudo service plexmediaserver stop
  1. Rename the Preferences.xml file to something else:
bryan@plex:/var/lib/plexmediaserver/Library/Application Support/Plex Media Server$ sudo mv Preferences.xml Preferences.xml.old
  1. Fire up PMS service and wait a few seconds:
sudo service plexmediaserver start
  1. Browse to plex, skip past configuring any new media libraries, and re-config any custom settings again.

I also added rebind-domain-ok=/plex.direct/ to my router’s dnsmasq config to allow dns rebinding. Now my server is working better than before. :slight_smile:

Thanks again to Chuck- big kudos for helping me figure this out so quickly.