Plex causing Synology to reboot? is this even possible? (long post)

Plex Info
Server Version#: 1.31.2.6783-6000
Player Version#: Various (Apple TV, Web etc)

Synology Info
Nas Model#: Synology DS3612xs & DS3611xs
DSM Version#: DSM 6.2.4-25556 Update 6
Both NAS boxes have been modded identically with upgraded CPU’s, RAM, 10gb NIC and Cooling, and have been ultra-stable for many, many years.
Both Nas boxes run a Plex server
Server ‘P’ for adult 12+ content,
Server ‘D’ for kids content (done this way due to the back-in-the-day lack of age restriction directly in plex), and both are nearly identical in configuration, with the exception of one box being the master store of media, sharing to both Plex servers, and the other server holding a backup of the master content, both boxes have 8x drives installed with approx 50Tb of storage. both. drive testing (simple and extensive, and data scrubbing) is performed on a scheduled basis

Issue
Over the past few weeks, one machine (server P) has been ‘shutting down unexpectedly’ overnight, this particular machine is kept in the garage, so figured this might have been caused by an environmental factor (temp etc) the error also suggested the power supply might be the cause.

What I’ve tried - Hardware (a lot)

  1. Figuring this was power related I swapped our the power supply for a spare I had for such an occasion, left for a few days, problem repeated. (concluded not PSU)
  2. I then removed 2 of the 4 Dimms figuring maybe Ram, run for a few days, problem repeated, I removed the remaining 2 Dimms and replaced with the 2 I had previously removed, problem repeated. (concluded not RAM)
  3. Removed NIC and ran system using onboard NIC, problem repeated (not NIC)
  4. replaced CPU cooler and ran from PSU directly not motherboard, problem repeated (not CPU cooler)
    • I have a replacement CPU on order to prove out the CPU, I doubt the motherboard is at fault for reasons I’ll explain later, however if it is, it’ll be nigh on impossible to source a replacement

What I’ve tried - Software (given it’s unlikely to be hardware)

  1. Stopped all additional packages from running on the Synology Server - RESULT! :wink: server didn’t crash
  2. Given Plex is likely the most intensive package running, I resumed all packages except plex, again all good, no crashes.
  3. Started Plex, crashing returned (it’s either Plex or the intensity of the app doing something)
  4. Assumed something was wrong with Plex, after much deliberation and caution, reluctantly removed plex (fragged the user, the shared folder and the package etc) and re-installed. all good no crashing.
  5. re-created all media libraries, crashed within about 20 mins
  6. recreated just the movie library, crashed within 20 mins.
  7. at this point I was beginning to suspect something to do with the drives, file structure perhaps or something relating to the media itself, I contemplated moving the hardware from server D to Server P, however as an easier step again reluctantly, I tried removing plex from Server D, and recreated the library (as I was trying to create on Server P) but this time on server D, Server D then crashed and rebooted about 20 mins into the library creation. (WTF now i have 2 NAS boxes crashing)

Still to try and conclusion
I have a replacement CPU on its way, and will try installing that, (but given step 7 above I no longer think this is hardware related.)
I’m going to try recreating the exact same libraries on server D as was there before, on the assumption it’s something in the adult folders causing the crash (highly unlikely but would help narrow the issue)
I’m going to try an older version of Plex (if I can find one) to see if the problem persists, if my guess that it is something in Plex doing this then i’d expect an older version to be fine (less this is something to do with the volume of media in someway or something corrupt/broken in the media folders)
Given the journey, it feels like something in Plex when scanning/creating the media folder is causing the issue. given the original overnight crashing I suspect a scheduled Plex task was doing something that caused the crashing, rebuilding the libraries during the day is effectively doing the same thing, just not scheduled.
I’m hesitant to suggest/consider that a package such as Plex can cause a server reboot, it suggests something is fundamentally wrong not just in Plex but also DSM in someway, I’ve not found anything that talks to a package causing a reboot of the underlying DSM/OS.

ASK
So you clever, kind, helpful fellow Plexers, what have I missed, is there something fundamental and obvious here going on that I’ve overlooked? have both my NAS boxes colluded in creating a niche hardware fault together? is there a setting here that can cause the hardware to reboot, perhaps? any ideas?

(where are the logs you ask)
yeah ok, I know, but just before doing that, can I ask what’s in the logs? i’d rather not be loading anything online that’s going to open up a raft of unscrupulous folks to ‘have a go’ if the content could be useful to that end in some way? I’m not against the sharing of the logs at all just looking for a little reassurance before I’ve got another challenge to solve.

Many thanks for scratching your heads on this one, it’s got me stumped, and contemplating a non-NAS future to avoid repeat.

The DS3611 / 3612 does not come with 10 GbE (predates it)

Is this an XPEnology box?

The 10Gb NICs are fitted to the expansion slots. boxes are DS3612/11xs

Have wiped plex and installed an earlier version (1.31.1.6733)

Created a new movie library, left Plex to do it’s thing.
few mins in, Nas box shutdown as before

Gutted, really hoped this was a version issue, going to try and find an older version and re-run the same test (still haven’t installed the replacement CPU)

ok removed 1.31.1.6733, and now installed 1.27.2.5929

ran exact same approach as before, but no crashing! Yay, great step forward.

I noticed that 1.31.1 (got to 70% at one point) seemed to hit the CPU harder than 1.27.2 (barely exceeds 40%) so I still have some doubts on the CPU etc. (I may revert to the original CPU to prove out the upgraded approach isn’t the cause)

however my current conclusion is something odd is going on inside plex that’s causing the issue. anyone else getting something similar?

Did you try turning off nightly credit detection? It’s been hard crashing windows and other OSes.

Going back to 1.27.2 would be the equivalent of disabling it as credit detection is a recent addition.

Thanks, DR_Tone, that would certainly explain what I’m seeing, (would have preferred not to have fragged my library to prove out, though)

If it’s been hard crashing other OS’s is there a thread to follow addressing this one?

Consider using 1.29.2.6364 available via Synology Package Center.

It is a very stable release. It does not have the problems with transcoding or audio analysis that plagued the early 1.30 releases. It predates Credit Detection as well.

I’m running it on my Ubuntu 22.04 system and not having any problems.

I would still inspect the CPU/Heatsink for overheating. The first release of Credit Detection certainly put the CPU to 100% for an extended time but unless other issues are present no crashes.

Given the age of the database, running a dump or recover on it wouldn’t be out of the question.

Thanks Tiebierius, i will check this out for sure, but given it’s the same result with 2 boxes i’d be very surprised if that’s the cause, particularly given the likely culprit of the credit detection. thus far I’ve not seen CPU temps higher than 51C, and that’s in a small hot (26c) room.

but your point is valid and I’ll certainly investigate further, though suspect it’s not the root cause on this one?

(@FordGuy61 yup will try that, thanks for the tip)

looks like this thread might be related (Plex Media Server Crashes Every night - #51 by Dr_Tone) talk about the credit detection causing crashes etc.

Folks,

I just got out of the support team meeting.

Engineering has fixed the Credit Detection feature breaking the host. Those changes are in our internal Alpha-testing with QA. Barring any breakage, IIUC, it will be in the next PlexPass release (please don’t quote me on that YET)

I would clean out the Syno (to be safe)
My DS1815+ gets dusty. I clean it once a year

Tangentially- regarding DB size / age, I have a tool which will clean up the DB to a large degree.
Please let me know if you want to use it.

1 Like

hello @ChuckPa thanks for the post, sounds like you have a similar approach to this as I do, an annual clean and check over etc, (oddly I find it therapeutic but maybe that’s my OCD kicking in.)

ref the DB size etc, given I’ve fragged all the Plex installs (shared folder, user, package) etc I don’t think I’ll have a problem, needing the clean up tool (less you tell me otherwise)

if you need any testing help on the new release more than happy to help, both my plex installs need re-doing anyway so don’t think I’ve anything to loose by offering.

QQ if i may do you know if there is a version number to look out for on this one, just so I’m sure which one to put the effort in the to plex re-build on please?

thanks

Just for giggles, i spotted a new version released a day or so ago (1.31.3.6819) so I gave that a shot, same reboots when building fresh library.

I’m seeing this crashing behavior, too. And I have disabled credit detection. I’ll have to look when I get home, but I think I’m up to 1.31.3.6819 now. This is NOT crashing the Synology NAS box, but the Plex service stops running, and it shows that it was “Manually Stopped”.

Was all of this crashing happening on the previous “official” version that the Synology Package Manager can install? I’m wondering if it would be a good idea to uninstall 1.31 and drop back to 1.29 until all of this gets sorted.

May I recommend backing down to 1.29.2.6364 (from Package Center) until things do get sorted.

Right now, I’m somewhat overloaded working on another problem.

As soon as that’s resolved, I’ll have time to come here and figure this out.

I am recommending 1.29.2.6364 because it’s the most stable version of PMS available to everyone. It is lacking some of the new features but, until verified, something in those new features might not be sitting well on these Syno boxes. That’s what I need to figure out.

EDIT: Posting the DEBUG logs ZIP file (manually created if needed … of the “Logs” directory) are of help

Sounds good. Is there anything I need to worry about with backing down? Do I just need to uninstall the current version and install 1.29.2.6364 from Package Manager? Will all of my recordings still be available after the downgrade?

You’re fine…

  1. Uninstall with KEEP option (Default & first choice)
  2. Install lower version
  3. Wait extra time (until CPU is completely idle) for database to revert to 1.29.2

Cool. I’ll give that a try. Thanks for the help.