Non-responsive after 1.15.0.659 ubuntu upgrade

[boomography]

Server Version#: 1.15.0.659
Player Version#: n/a

Attempted an upgrade to 1.15.0.659, appeared to go smoothly, but server is unaccessible upon completion. Attempted to repair databse, just in case, still no joy. Attempted to downgrade to public 1.14.1.5488 still no joy.

Taking a look at the Plex Media Server logs, it looks like the last time it logged was before I attempted the first upgrade to 1.15.0.659, it doesn’t look like anything is being logged past that point. Journal shows plex service starting up. I tried renaming the db and dropping the original to force plex to create a new one, and still no joy.

Have also tried removing the install via apt-get and re-installing both public and beta release, to no joy.

Unsure what I should be trying next. Logs attached.

plexlogs.tar (3.1 MB)

[casedog21]

I am seeing a similar issue.
Plex is bring my linux box to its knees. The last two versions are showing a memory leak.

What does your ram usage look like?

This is mine while plex is running.

This is with it stopped.

-Casey

[casedog21]

Not sure my issue is the same but I took your method and restored an older db and I am buck up and running again as it should.

[hstorey219]

I am running ubuntu 18.4 headless and the Plex version 115 has a memory leak, using 8 gig of memory, went back to 114 and now at 600 meg.

[thesav]

I seem to be having a really similar issue - how do I go about restoring an old db or downgrading the version?
Thanks

[ChuckPa]

Need logs please and how you caused this to happen…

I will do my best to recreate in VM and backtrace

[hstorey219]

Chuck, How do you want me to send the logs,? I can send a link. I just reinstalled 115, give me short time to get the logs and, under 2 mins and up to 2 gigs already. 114 never went above 800 meg

[hstorey219]

Thesav,

I just download the version 114 and do dpkg -i command . It will downgrade to this version, I always do a reboot, just incase

[ChuckPa]

If you’ve re-installed 1.14 there is no further point. I

I will have to wait until someone has live data.
Your logs will already have been superceded

[hstorey219]

Plex Media Server Logs_2019-02-13_01-02-01.zip (2.0 MB)

[hstorey219]

Chuck,
This is from the version 115, I reinstalled, after 10 mins around 1.5 gig already. I can leave it run over night if you need more data

[ChuckPa]

Ok… already seeing a bunch of things.

I have run 1.15 for days and not seen this.
What I see in your logs is DNS rebinding problems / plex.direct is not declared as a valid private domain.

Feb 13, 2019 00:48:25.725 [0x7f98a8ff9700] DEBUG - Connection Tester: Successfully tested media server FusePlex to be at https://68-255-238-179.e33c2fdaaf084d2faba1e2bee7ca6714.plex.direct:32400 in 118 ms.
Feb 13, 2019 00:48:25.765 [0x7f986bfff700] ERROR - Error issuing curl_easy_perform(handle): 60
Feb 13, 2019 00:48:25.765 [0x7f986bfff700] WARN - HTTP error requesting GET https://plex.keegami.net:32400 (0, No error) (SSL certificate problem: unable to get local issuer certificate)
Feb 13, 2019 00:48:25.766 [0x7f9961ffb700] DEBUG - EPG[onconnect]: Matched channel 035 () with grabber channel, adding.
Feb 13, 2019 00:48:25.807 [0x7f98a9ffb700] DEBUG - EventSource: Resolving 47-185-129-102.233d5cc0f068488b9a41bf1e70916b47.plex.direct port 32400s
Feb 13, 2019 00:48:25.807 [0x7f98a9ffb700] DEBUG - HTTP requesting GET https://47-185-129-102.233d5cc0f068488b9a41bf1e70916b47.plex.direct:32400/media/providers?X-Plex-Token=xxxxxxxxxxxxxxxxxxxx
Feb 13, 2019 00:48:25.807 [0x7f997a697700] DEBUG - EventSource: Resolved 47-185-129-102.233d5cc0f068488b9a41bf1e70916b47.plex.direct to 47.185.129.102
Feb 13, 2019 00:48:25.925 [0x7f98f0ff9700] DEBUG - Request: [127.0.0.1:54506 (Loopback)] GET /system/messaging/clear_events/com.plexapp.agents.movieposterdb (8

[hstorey219]

Hmm, so I am wondering what is changed from 114 to 115, I have not touch my Pfsense router. 114 doesn’t have this problem, I can reinstall and it takes no more than 600-800 megs.

[ChuckPa]

My config

[hstorey219]

I have the same except for the so-reuseport: no

[ChuckPa]

Time to spin up the VM again

[hstorey219]

Sorry

[hstorey219]

Here is the htop snapshot

[ChuckPa]

It’s probably fallen into a hole trying to get the cert. Look at the logs please.

Do you have PMS customized in any way?

[hstorey219]

I have HomeRun for live TV that is it. Noticed how many instances are running, is that normal?