I just updated the plex to plexmediaserver_1.18.4.2171-ac2afe5f8_amd64.deb. A straight install over the existing version. Did a ‘systemctl daemon-reload’ and restarted the service.
Now my server is completely unreachable. This is the second time this has happened. The last time a fix came about ~30 minutes later. Can some get his correct the first time? Is anyone else experiencing this? Is a fix in place? Please advise.
So I update my PMS using a script I wrote (b/c somehow I hosed the initial install and have to replace the user/group with ‘root’, don’t ask). My script has worked great: install the new *.deb, reload the daemon, restart the plex service, delete the *.deb, done. It has worked perfectly at least a dozen times. I would just open the web page and verify the new version was installed.
Now, it has stopped working. The script completes and the service is running fine, but when I open the web page, all of my libraries are missing. Somehow I log out and forget the server, then I can get it to run the setup all over again, and it finds all my libraries/metadata. The real PIA is that I have to go around and re-register every device.
Thank you for your kind words. I was truly out of line.
I was not able to identify the problem, unfortunately. I’m not sure if my update script was causing the issue, or if it was something else. I have used the script many times before with great success. But the last two updates, released from Plex, did not go so well. Again, no idea where the fault lies.
My script would install the new *.deb file first. My need for the script stems from the fact that I made some mistake in the initial install, which executes the PMS scripts with the user ‘plexms’ (my main user on the linux box). I am dangerous with linux, not an expert. So I know that I have a permissions issue somewhere. A PMS update breaks the install, and I need to edit the service file to have ‘User=root’ and ‘Group=root’, daemon-reload, service restart (all done by the script). After that, all is back to normal (usually).
That worked many times before, but as I said, it caused a major headache last two updates. After the update, I open the PMS web page and it can no longer find my PMS instance. I eventually have to remove the old server from the ‘Authorized Devices’ list, and then rerun the PMS setup. It finds my PMS instance and all the libraries. I then have to fix my settings and re-authorize all my clients. It’s not really that bad, it’s just time consuming.
There is a problem with the software deployment mechanism right now. The team which handles it is working on it. I suspect the problem seen above is related because the last two deployments by automation were problematic. This coincides with above.
I wish to advise of an upcoming change to the installation scripts in the Debian packaging. It will not impact any scripts which use the return codes ($? / $status) variables. The changes will impact if the package installation output is parsed.
As reference: please look here: New Linux installer - Forum preview & beta testing
Having looked back through all my notes; the actual installation itself has no bearing on being able for Plex/web to find and open the server instance.
The bundled Plex/web did change.
I have had reports on other distros (NAS platforms mostly) where logging in again was required.
It’s possible this happened here too; the server had become disassociated