The reason I like docker based Plex is because it can auto-restart on crashes with ease. But I need HW transcoding so I am going to stick with the Synology package.
So want to know of a way to have it auto restart on crashing. Thanks.
How do you define “crash”? PMS processes are no longer running (traditional Linux ‘crash’) or “stops responding” ?
If you’re talking about either, one is more easily done than the other but you need to be willing to a) learn how Synology does package management and b) write some shell script. Neither is difficult but you will have to get your hands dirty if you want it automated.
If you want to restart, Package Center is already in place to let you stop / start using a GUI and that’s pretty easy.
PMS crashes and process no longer running. I can start again from GUI but sometimes I’m not home and still want family to be able to use Plex.
PMS shouldn’t be doing that especially on Synology. I run it on Syno, and on QNAP. It runs for weeks without any incident. The only time it stops for the bulk of us is when we update to a new version.
How about we try to solve the problem rather than find a work around?
If you concur, the next time you restart it, Immediately go to “Settings - Server - Help - Download Logs” and attach the ZIP file it gives you here. We’ll find out why it actually crashed.
Sure, I’ll do that next time. I think the problem last time was cpu usage sky rocketing when comskip was working. I think.
Same problem here…
i use synology DS916+, latest DSM 6.1.5-15254,
and plex version 1.11.3.4803
my plex is crashing…meaning that it stops working…!
so if i try to stop it and relaunch… it doesn’t work.; it seems the only way is to restart my synology server
i attach a pic of the log… maybe it is because of this error it stops functionning
thank you both… i will try
I’ve been running my server on my DS216j for about a week now and I’ve had to manually restart PMS twice. I’ve attached my logs, so analysis would be appreciated, thanks.
NJMorf,
Your internet connection from the Synology is not finding the internet.
Plex Media Server.2.log:Feb 16, 2018 21:25:48.446 [0x7f9e49559700] WARN - HTTP error requesting GET https://plex.tv/updater/products/1/check.xml?build=linux-ubuntu-x86_64&channel=0&distribution=synology&version=1.11.3.4803-c40bba82e (0, No error) (Couldn't resolve host 'plex.tv')
Check out Control Panel -> Network -> Interfaces
Make certain the DNS server is correctly configured to point to your modem/router or it’s using DHCP (where the modem/router provides the configuraton)
My NAS is plugged directly into the modem. I’m using the NAS as the DHCP server and it’s pointing to the same DNS servers as the modem does (OpenDNS servers) and the modem’s IP is set as the gateway. I’ve added the DNS server IP to the Network Interfaces page (it wasn’t set on that page, I don’t know why). I’ve also been using fixed IP addresses for most devices on my LAN.
Is PMS really crashing/exiting because it can’t find the internet? That seems like something that shouldn’t be fatal.
- PMS requires the internet to download all codecs
- PMS requires the internet to validate your Plex Pass.
- PMS is not written to be run ‘offline’ (disconnected).
It’s one thing to not be written to run offline, it’s another thing entirely to crash or close the server without notice just because you can’t reach the internet. Doesn’t it just sit there and wait for the connection to come back up?
Regardless, based on your answer I can only assume this is working as intended, so we’re back to the original question: is there any way to automatically restart the server when it crashes/closes?
PMS stopped/crashed again today and I’ve only just spotted it. Third time in seven days.
According to DSM, Plex already has read/write access ( in both the preview and group permission columns) to the entire Backups share. I’ve changed the permissions as detailed in your link, but it hasn’t (yet) made a difference when looking in the Control Panel - Shared Folder app.
As for recordings being in the past: when I moved the server, I recreated all of the recordings that I could from the freshly-downloaded EPG, so it’s unreasonable to assume that there are any, or at least significant numbers of, upcoming recordings that are actually in the past. I’d already set the programmed recording for the show you flagged to disallow partial recordings - is that error just saying that it won’t record this program because it’s a partial? In the latest logs I’ve pulled from the server a few minutes ago, there are only two instances of that error message, both for the same episode of the same program, 13 seconds apart. The server does not appear to have crashed out in response to either message, based on the log file.
The server crapped out again, probably sometime between about 05:30 and 09:00. Something, almost certainly PMS performing overnight maintenance, was using significant CPU between 02:00 and 05:00ish (the times I’ve set for jobs to run). The logs I pulled only go back as far as 13:25, about the time I manually restarted PMS, so they’re no use at all.
I’m running that check on my video files, but given that it has to transfer them all from the NAS to my desktop, it’ll take hours. So far it has found a few files with errors (mainly frame sync errors and incomplete frames, which sound fairly innocuous to me), but none that have been recently recorded or watched, except when I just tested them (without any evident problem). The only reason I can think of for them causing a problem would be if the video thumbnail generation balked at them.
@trumpy81 said:
You obviously did not read the FAQ’s correctly. There is no need to transfer the files at all, you can use the UNC address to check them, although that will be a little slower due to the network connection etc.
I read this in the FAQ:
If you have a lot of files in your path then this batch file will take a long time to complete as each file will be read and filtered through ffmpeg. Each file will take approx. 3 or 4 minutes.
Given that the process is running on my desktop, not my NAS (unless I’m missing something else), and given the amount of network activity I saw after starting the task, it was (and is) a reasonable assumption that the data is being transferred from the NAS to the PC in order to check it on the PC. I didn’t mean that I’d manually copied the files across.
That said, it never manages to check more than about 10-15 videos before the batch fails for some reason, assuming it starts at all (I pointed it at deeper subfolders and it failed to check anything within them, without adding to or creating a log file to explain why).
Permissions seem to be behaving themselves now - I couldn’t find any errors related to the folders Plex didn’t seem to have access to before.
PMS hasn’t crashed for about a week. Maybe all this was teething troubles related to the workload of processing all that data from scratch.
Network problems are, I’m fairly sure, entirely down to the flaky nature of the powerline ethernet. The NAS does have a static IP: the NAS is the DHCP server on the network (and DHCP is disabled on the router). I still maintain that crashing PMS because the network/internet is missing, or even deliberately shutting it down, would be an incredibly poor design choice and fixing it should be very high priority.
