QTS 5.0 5.0.0.1891 and Plex Server Hangs

Server Version#: 1.25.5.5492
Player Version#: Various

I finally waited until last month to upgrade to QTS 5.0 watching the release notes and ensuring most things would seem to be stable. Everything has been fine in fact with one odd exception.

The Plex app stops responding randomly but on a semi-regular basis. I need to go stop the app and start it again for players to see it. I even have some monitoring tools that tell me when the Plex Server port is no longer responding. Can someone point me where to look in the logs as to why it is just stops responding but an app restart “wakes” it back up again?

I can say this was never the case with QTS 4.x and there were numerous threads during the QTS 5 BETA about the plex server app, but most appeared to be addressed and resolved. QTS 5 has been out in GA form now since November I believe.

Thanks in Advance!

I am also having this issue with QTS 5.0.0.1932 and Plex Version 1.25.5.5492

Thank you for that report. I was thinking about upgrading my QNAP firmware to v5, but I just re-considered.

I don’t want my PMS service to become unstable.
Please, @colotti and @ben1103 if a Plex ninja/dev/representative pops up here, would you please so kind to provide him or her your logs in order to assist them fixing the issue or find the new firmware setting that keeps PMS from running stable?

@ChuckPa any insights?

This is kind of Important. The lateset 5.0 FW fixed some of the Deadbolt vulnerabilities. Might be safest to go back to the last 4.5 FW which is stable and was also patched for Deadbolt

Yes if he happy to provide any and all
Logs as directed to do so to help get the issue solved. We just need someone to say they can look into it and which logs they need. I suspect some Plex logs and qnap logs might be required but I’m not sure.

More than happy to provide data though.

I’d say eventually They will want everyone to move to QTs 5 in 2022 and again it’s currently more of an annoyance but it’s “randomly consistent”. Meaning it stops responding randomly but it’s consistently happening over time and nothing more than an all stop start is needed.

Now this doesn’t mean there is T something bigger at play under the covers though this is just the symptom.

I am sure QNAP will but for now 5.0 still has A LOT of known issues and they have even added more, really 5.0 is an unfished work. I do not consider it production stable. I know QNAP does :roll_eyes:

But apparently they are self aware enough that they “back patched” the last 4.5 FW version for Deadbolt, so I suspect internally they know 5.0 is not quite there yet.

@skwor01 True enough, but the catch-22 is if we all just roll back then we will still eventually see the issue so for those of us willing to try and help at least we can do so. I could roll back too, but it’s just as important to provide the data to the Plex folks to see about resolving it for later users who eventually upgrade. It’s definitely much better than it was in 2021 (QTS 5 that is). All my Docker stuff has been fine so far this minor plex issue is to date the only app that’s appearing to act up a little bit.

TO ALL:

Let me read back and catch up :nerd_face:

Updating my QNAP to latest firmware.
Also will install the latest PMS because it contains the scripting changes I implemented for QTS 5.0

Will post again shortly.

@ChuckPa thanks! Let me know how I can help. I can check my monitors and give you a MTBF on the monitors of the service but it is in some cases a few days or more between the hangs. Give me a few and I’ll post those time stamps before I head out of the house after lunch.

My TS-128 is working ok (still). :sunglasses:

I want to advise everyone, PMS 1.25.6 contains the scripting updates I wrote about a month ago (sorry about the delay getting them released).

These are the ones which fix the App Center control problem (PMS hangs /fails and a reboot of the NAS is needed)

I advise updating to that

Updating my TVS-1282 now

@ChuckPa is 1.25.6 BETA/Pass Channel only still? I assume so but when will it be the public version? Just curious I have rarely used the Beta channel over the last 3 years mostly public, but if in this case it makes sense to do so as a test I can.

FWIW here is the outages recorded on mine

|Down|2/15/22 9:43|Connection Timeout|
|Down|2/14/22 23:50|Connection Timeout|
|Down|2/14/22 10:26|Connection Timeout|
|Down|2/7/22 20:45|Connection Timeout|
|Down|2/6/22 9:11|Connection Timeout|
|Down|2/4/22 7:29|Connection Timeout|
|Down|2/4/22 7:24|Connection Timeout|
|Down|2/4/22 7:14|Connection Timeout|
|Down|2/3/22 20:03|Connection Timeout|
|Down|1/28/22 15:36|Connection Timeout|
|Down|1/28/22 15:20|Connection Timeout|

Yes, It is beta channel
There are a lot of fixes in this version. I feel it’s worth it

Question about those outages –

Have you contemplated upscaling the script to calculate duration?
( a counter in a loop would do it )

I have the duration down from uptime Robot, but it’s basically relative to when I get the alert and can restart the app. SO duration is…sorta…Meh LOL.

Uptime robot is just monitoring the external access to plex but then I confirm it’s non responsive internally as well and do the restart. By then duration could be 5 mins or hours if it hangs overnight. See below so it’s not a great data point since the app restart is manual intervention

Event Date-Time Reason Duration
Down 2/15/22 9:43 Connection Timeout 1 hrs, 25 mins
Down 2/14/22 23:50 Connection Timeout 0 hrs, 1 mins
Down 2/14/22 10:26 Connection Timeout 0 hrs, 5 mins
Down 2/7/22 20:45 Connection Timeout 0 hrs, 7 mins
Down 2/6/22 9:11 Connection Timeout 0 hrs, 9 mins
Down 2/4/22 7:29 Connection Timeout 0 hrs, 2 mins
Down 2/4/22 7:24 Connection Timeout 0 hrs, 4 mins
Down 2/4/22 7:14 Connection Timeout 0 hrs, 9 mins
Down 2/3/22 20:03 Connection Timeout 0 hrs, 3 mins
Down 1/28/22 15:36 Connection Timeout 0 hrs, 8 mins
Down 1/28/22 15:20 Connection Timeout 0 hrs, 9 mins

I will pull the Beta channel version and try it. Do you know when that will get pushed to public anyhow? so I can re-converge back to Public at some point?

It usually only takes 2 weeks but if Engineering has a reason, they will push sooner.

Also, feedback from us about the forums has a lot of impact.

If you need help with that scripting, I’ll help.
crontab would be a nice way to restart :slight_smile:

We can also add some metrics to those outages (see how busy it was at last sampling)

@ChuckPa yeah you don’t want me writing any scripts… LOL. There are some webhooks that can be sent from Uptime Robot that could be tied in to restart the app, but let me get the beta version loaded now and see how that rolls first.

Version 1.25.6.5545 installed now so let’s monitor and see.

First thing you’ll notice. Start/Stop WORKS :slight_smile:

@ChuckPa That’s good! I will say I had not yet had an issue with the app center stop/start (yet) it seemed to work each time the app stopped responding so that’s interesting. Let’s see it I get any down alerts over the next few days / week or so.

The specific situation which failed –

  1. PMS would crash.
  2. EAE would be left behind.
  3. PMS would attempt to start but EAE still held the connection to 32400 open (was talking to it)
  4. Start failed.
  5. Stop failed because PMS was already stopped.
  6. REBOOT required :rage: