Everything looks good as far as I can see! 
So you can’t break it / get it locked up / make the App Center controls fail to put PMS back to right ?
So I should do that?
I’m making a backup of the Plex metadata files now as explained in the FAQ…or at least in a similar way: I’m doing it using Windows Network where I can see the Plex Media Server folder located on my Qnap, and now I’m compressing the 8.5 GB of folder I have (btw…is it normal to be this big?!)
After this 1h that it will take I could try to restart the Qnap totally…should I give it a try?
In case, should I first restart back on PMS first??
Thank you.
Steve,
Is PMS still running the update ?
The reason for the LONG startup is because of a major database change
not so far, but by all means give me some things to try that I might not have thought of.
J.
the worst you can do it it would be randomly kill certain processes.
( try to emulate pieces of PMS failing ).
The most dramatic failure is when PMS itself fails (crashes) and fails to shutdown all its slave processes. That’s what this update focuses on – making certain that never happens again when you click “Stop” in App Center.
We all know how painful it is to restart QTS
Yes, I’m still with 1.25 and after a full night I can’t access my PMS web page…was it still doing its major database stuff? Should I leave it longer?
Steve,
Does the CPU show as idle ? It should not be doing anything at this point.
The database update should require no more than 15 minutes .
Go ahead , create the PlexData shared folder, Stop/Start Plex (so PlexData is populated) and retrieve the logs.
We’ll find out what’s going on.
I suspect either a) DNS/gateway or b) it got signed-out of your account.
Let me finish the Plex Media Server folder backup. It is archiving the full 8.5 GB folder into a ZIP.
Once ended, as Plex is still stopped, I will extract the 1.25 Crash Report folder. Need anything else?
Steve,
Before sending, Please confirm the Crash Reports folder has DMP files in it.
If it doesn’t then I there is nothing for me to look at.
I open the DMP file in the debugger and see where PMS stopped (the line of code)
Ah!! You are right: the folder is empty.
How to populate it with the .DMP files?
Settings - Server - General - Show Advanced.
UNCHECK the box – “Send Crash Reports to Plex”
SAVE.
The next time it crashes, the DMP file will be in the directory
I have no custom scripts running. Regarding apps, almost all the apps I have in my QNAP are “official”, with exception of Syncthing and Radicale.
PMS stopped working after a QTS upgrade, though.
After a Qnap restart, I’m able to access the Plex web interface.
And now I’m able to reach the “Send crash reports to Plex” checkbox (unreachable before)…and it is already unchecked.
Btw, I see a new version of PMS. Should we install it safely?
Just came here to say my Plex has been doing this for a few weeks quite randomly, but since the latest update to QTS 5.0 (and I updated plex to the latest as well), this happens daily. I am not so experienced in SSH/Linux file system but if you need more testers eventually I would be willing with explicit instructions too =).
Is there a recommended backup method via SSH that anyone can recommend (and itemize the steps possibly)? THANKS!
- Do not update yet. 1.25.2 will contain the new control script I’ve just updated. I prefer you start with that version. I rewrote the segment which handles unexpected crashes/failures to be 100% certain Plex cleans up after itself when you click ‘Stop’ in App Center.
You can make a backup via SSH or via FileStaton
Using FileStation, use the PlexData shared folder
For both cases, you want to get into the “Plex Media Server” folder, with Plex stopped, and then make the ZIP or tar.gz of all which is there.
1 Like
Folks:
I have PMS 1.25.2 in hand. I’ve tested the updated control script (plex.sh) and it works as it should.
Given this package hasn’t been fully QA’d yet (it’s considered ‘alpha’ stage), is there anyone who’d like to give it a test drive ?
If so, please tell me the CPU architecture when you make your request.
1 Like
To add to @dane22
This update includes a process control improvement which better handles those cases when PMS stops abnormally and leaves the package in a bad state.
It is now more prudent about stopping the package.
Also, the control script does not look for GPUs unless the NAS is an x86_64 platform .
I’ll give it a shot, but it will be tomorrow now (I am in the UK).
