Same here. My console is clean now, just the expected stuff no looping commands & things seem to be running better
@ChuckPa First crash with 4276. Things started off very smooth but the longer the server was running, it slowly started getting very sluggish and eventually crashed.
A few days ago I went back to 4136 until things settled down a bit, then yesterday afternoon I updated to 4726 when 4136 crashed for the first time.
My resource usage has been normal the whole time.
I donāt understand why itās crashing, Iāve gone back to 1.21.2.3841 and I still get crashes. Iām optimizing the database and cleaning bundles now, resource usage is still normal.
1 Like
Bouncing back and forth between versions when thereās database changes occurring is probably not the best for stability. Iāve been stable for the last ~3 versions but Iāve not down a downgrade.
I did mention that up before, but I didnāt hear anything about it so I figured it wouldnāt be an issue.
Iād also expect to see something in the logs about it, I havenāt seen any indicators for why itās crashing.
At this point I was on 3841 and have checked the database for corruption, came back okay and have gone through the database repair process (just because). Iāve then updated back to 4275 and letting the server come back completely on its own. See if it handles that any better.
Iām hoping it was just something nasty as a result of flopping between versions during that mess earlier this week.
I ended up having to grab a āsaneā database version from a backup from before the database changes and allowing the migration to happen with that copy. After doing this and merging over user watched settings everything seems to be functioning normally for the moment.
With the changes to the database, the documentation for anything database related I think should be updated.
My one and only indicator that my issue was a database issue was I noticed any database changes were progressively taking longer and longer to finish in the logs.
Is the backup something you did or plex does automatically?
My server is crashing every few hours after the same updates mentioned in this thread. Some have said they have mixed results will rolling back the version, is it possible to roll back after the schema change?
Thanks!
CentOS Linux release 7.9.2009 (Core)
PMS 1.22.1.4275
Plex does have a database backup feature that can run every 3 days, but this wasnāt enough for me since Iāve been fighting with this for 2 weeks.
I run my own backups and had to grab a version from a month ago and allow the database change to happen on that.
For some reason Iām back to the same looping
XML: Entity: Line 1[]
XML: parser[]
XML: error:[]
XML: Start tag expected, '<' not found[]
XML: status=ok[]
XML: ^[]
Error parsing content.
Error parsing XML: Error parsing file.
SSDP: Error parsing device schema for http://192.168.X.XXX:9080
Everything has been getting slow & I had new stuff added that wasnāt showing until I stopped & restarted the server app.
But I noticed something else this time. Thereās no devices playing anything, the dashboard is clear but after the repeating reds thereās a bunch of non-flagged text that seems odd
Like Request: [127.0.0.1:62962 (loopback) GET /library/metadata & We're going to try to auto-select an audio stream for account 1. & We're going to auto-select a subtitle. & Audio Stream: 219417, Subtitle Stream: - 1 that keeps repeating only the Audio Stream:#### changes, but nothing needs an audio or subtitle stream so what is it doing? It never used to do this
How did you merge over the watched status for everything after the backup you restored?
1 Like
I used the instructions found here:
The process does appear to still work, but I feel like all of Plexās current documentation is now out dated due to the database changes. After running the commands I got an error about things missing, but it doesnāt appear to have affected the end result.
Hi
I am having the same issue stated in this thread. My Plex Server crashing after every few hours. I made two changes to my Server, first i did the update of the plex version stated in this thread and 2nd I upgraded my Storage Unit so now i am confused what is causing the issue.
I would really appreciate if someone can guide me to resolve this issue. Attached are the logs.
Thanks.
TZPlex Media Server Logs_2021-04-06_22-04-39.zip (4.8 MB)
What ISP do you have if any? The :9080 issue (at least for me) is caused by Plex using some service discovery protocol and assuming XML responses, however, my ISP (Optimum) uses that port on itsā STBs to assist in authenticating to Netflix, and if you visit that ip/port in a browser you just get a āstatus=okā message. If you open netflix on the STB and log in, the port closes until the auth token needs to be refreshed, and as a result youāll momentarily not have this in your logs.
This is unrelated to the issue with Plex crashing, but Iām wondering if this is a new Optimum thing or a new Plex-assuming-XML thing.
The only solution Iāve seen is to hope you have a backup of your database from before you updated, start Plex using that database - if youāre on 4275 or newer the database migration will occur on startup, the web UI wonāt become available until the migration is complete (be patient),
then youāre done.
It appears Plex did make a backup a couple days before I upgraded to 1.22.1.4228 (and then 1.22.1.4275).
I see a couple of dated db files: com.plexapp.plugins.library.blobs.db-2021-03-25 & com.plexapp.plugins.library.db-2021-03-25
Can I do this?
- Shutdown PMS
- Backup current db files
- Rename backups from 2021-03-25 to production
- Downgrade to 1.22.0.4163 (version when the backups occurred)
- Start up PMS
Seem reasonable? Flawed approach?
Thanks!
I see a new version got published, with mention of a fix for a ādeadlockā. That sounds promising and will try that before taking a more extreme approach.
Nailed it! Thatās exactly how you do it. Donāt downgrade though, the database will migrate on next startup of Plex, just be patient.
Keep your server on the version itās on now, rename one of your backup dbās to production, start Plex and just wait. The web UI will become available after the migration is complete. On my dual xeon server it took about 6 minutes to complete the migration - your mileage may vary.
The only thing that seems to keep happening now in logs that I notice after it becomes unavailable are these:
pr 08, 2021 09:00:03.497 [15752] ERROR - NAT: PMP::getPublicIP, couldnāt initialize: -3:Unable to get Gateway
Apr 08, 2021 09:00:06.508 [15752] ERROR - Error issuing curl_easy_perform(handle): 28
I didnāt change anything with my network setup here though, and I donāt understand why anything like this would stop the server from being available even on the same machine (Plex client on Shield canāt see the server also running on Shield it when it happens) and I canāt even access the server by IP on the LAN when it happens either.
Does anybody have any idea what exactly itās doing here with these curl errors?
Follow-up. After upgrading to 1.22.2.4282, I have been stable for 4+ days. I did not pursue a database rollback.
1 Like
Did a sever update and once it was completed it could not reconnect. when I double click the plex media server desktop app it just gives a pop up saying error. I am connected to the internet and have no idea what is wrong and why it wont connect to the server. Please can you assist.
