Have a Synology DS1815+ running a fresh install of Plex server 1.13.5. Suddenly it’s offline after a day or so of working fine. I’ve made no system changes other than to be adding my movies to the single library I have configured at the moment.
Can logon to the servers IP address but even connected there the server shows unavailable.
I’ve tried reinstalling and no dice. It was working great then all of a sudden…nothing.
Here’s what I see both connected directly to the NAS as well as at plex.tv
I’ve restarted the NAS several times as well as stopping and starting it. This has happened multiple times as I’ve reinstalled 3 times now. Every time it works for a while and then suddenly stops.
Go to Control Panel - Shared Folders - Edit the Plex share – Permissions – Give your username R/W permission to it.
Open FileStation, Navigate Plex / Library / Application Support / Plex Media Server
Right-click Logs , Compress to Logs.zip
Attach that file here with your next post. I’ll look and see what it’s reporting.
Shared folder permissions for Plex… I went ahead and pushed the permissions down to the subfolders and files just to be sure.
After another restart I could load the server and my libraries, but trying to edit tags, or play anything times out and I get a message like “error opening item” or something. I’d add the specific error here but it’s offline again.
Logs.zip (2.6 MB)
Instead of accessing it through Plex.tv, open it directly http://10.0.0.5:32400/web
If your computer is not on 10.0.0.x subnet or you are remote please tell me
Following this thread as I’m experiencing a very similar issue as of yesterday. I’m on a QNAP TS-251+ so this issue may not be NAS specific. It’s suddenly acting the Plex Media Server stopped broadcasting.
Attach some logs and go from there?
If you see in the logs for this thread,
Aug 10, 2018 10:12:14.355 [0x7f7097ce3800] DEBUG - Network interfaces:
Aug 10, 2018 10:12:14.355 [0x7f7097ce3800] DEBUG - * 1 lo (127.0.0.1) (loopback: 1)
Aug 10, 2018 10:12:14.355 [0x7f7097ce3800] DEBUG - * 7 bond0 (10.0.0.5) (loopback: 0)
Aug 10, 2018 10:12:14.355 [0x7f7097ce3800] DEBUG - * 8 docker0 (172.17.0.1) (loopback: 0)
Aug 10, 2018 10:12:14.355 [0x7f7097ce3800] DEBUG - Creating NetworkServices singleton.
10.0.0.5 is the IP address PMS is looking for. This is the IP the host is reporting and should be accessed with.
You can look at your Network config in QTS and find the same information (vswitch or otherwise)
I experience the same behavior accessing it from plex.tv, the 10.0.0.5 address from my local subnet, and directly to the external IP of my router where I have port 32400 successfully forwarded.
I’m currently remote, but have TeamViewer access back to my house for local testing.
In addition to Plex I also have Docker with a Ubiquiti AP. This as well as all other functions of the NAS seem to be operating normally. PLENTY of available resources.
Is PMS running IN Docker? If so, everything I have been working toward is now changed.
There are only two things I need to know upfront ( VM? Docker?)
No, plex is installed as a package, I was just saying, other stuff is working so the NAS itself is not having issues.
Thank you for being clear. It’s been one of those weeks and I apologize if i’m a bit ‘crispy’ around the edges 
A big change in PMS 1.13.5 is the ability to differentiate which ethernet adapter PMS binds (listens) to.
It’s been needed for us on Synology and QNAP a long time.
What many people are now seeing is, because of this change, PMS will gravitate to the docker interface for communication until you expressly select one of the physical DSM interfaces (eth0 bond0, etc)
My DS1815+ is configured with host IP assigned at eth0. All four adapters are active. I’ve set PMS to bind to eth0. There are no connection problems.
I suspect this is all you need to do.
Settings - Server - Network - Show advanced. You’ll see it there in the drop down
No problem, I wasn’t clear.
I have all 4 interfaces bonded into a single connection that has the 10.0.0.5 IP address. When I go to settings there is no server option with the server offline or unavailable. Is there a way to edit a config file or something to change this setting and restart the service maybe?
Now the server isn’t responding at all on the local port… FML.
@ChuckPa that definitely sounds like a strong lead for what I have going on but I don’t think I was on 1.13.5 until I upgraded this morning trying to resolve the issue.
I’m not quite sure how to “set PMS to bind to eth0” as you stated. Is that from within PMS settings? If so I can’t even seem to access this specific PMS. Just other external PMS servers.
If its a setting from within my NAS, I’m also having issues tracking that setting down. I realize you have a Synology but I can’t seem to find any specific way to bind PMS to eth0. I looked in the settings and it looks like there is a binding option that was unchecked but when enabled all options (none look app specific) are already checked and when saved…it doesn’t seem to resolve the issue.
I believe he said you do that from within the PMS server settings not on the NAS.
Yes, there is a manual entry for Preferences.xml
QNAP bonded interface shows as: qvs0 (the bond is behind the virtual switch) .
Synology bonded shows as bond0
For you, the Preference to add is ServerBindInterface="bond0"
Save and restart PMS
Yeah but see that’s my issue. I cannot see the server at all from within the web interface that is functioning on the very same IP address (subnet and all) as what PMS should be working from. It’s as if the PMS has simply stopped announcing itself. It can be found from within the processes but I can’t see it in the web interface nor my mobile device on the same network. I would think i should find it on the same local IP right? It’s like its in stealth mode. That’s why this whole “bonding” thing sounded promising but now I’m wondering if my issue is more complicated (or idiotically simple) than that.
I did the same as @JaMike_Richardson and edited the preference.xml to add ServerBindInterface=“qvs0”. Saved, restarted PMS and still no changes. 
So, what, just add that to the end of the XML like so?? If yes, it didn’t seem to change anything.
<Preferences (yada yada yada) PubSubServer="184.105.148.102" PubSubServerRegion="sjc" ServerBindInterface="bond0"/>
That’s it’. Now PMS will listen only on the IP address of “bond0”.
To verify PMS is alive and you have connectivity to it, linux curl is perfect.
[chuck@lizum plexmediaserver.117]$ curl http://192.168.0.21:32400
<html><head><script>window.location = window.location.href.match(/(^.+\/)[^\/]*$/)[1] + 'web/index.html';</script><title>Unauthorized</title></head><body><h1>401 Unauthorized</h1></body></html>[chuck@lizum plexmediaserver.118]$
Notice the ‘Unauthorized’ reply in HTML format.
You should have the same basic capability in a browser:
Drop the /web from the end and check the reply.
OK… After the change and restart of Plex I saw no change. I stopped the package service, restarted the NAS and the started the package service again and I’m in. I can play and edit titles again.
Lets hope it lasts, it worked for days before this happened.
Thanks SO MUCH for the help, and so prompt. I’m impressed!
