Unraid 6.8.3 slows to a halt a few times an hour

[gregtamaki]

Hi,

 

I have been having an issue over the past 2 weeks where everything about my unraid setup (mostly used for plex and the related binhex sonarr/radarr/jackett/delugeVPN dockers) crawls to an unusable halt for anywhere from 5 to 15 minutes regularly. Things were fine for months with a handful of users. Previously, I was at about 3 months uptime and kept all dockers up to date etc. I have clean shutdown when having issues and stopped different dockers to see if that was it, but none of those seem to be the issue.

 

By locking up, the following things happen:

-Plex streams stop, menus don't load on connected devices

-UI won't load for the dashboard or any docker UIs I check

-Control IR plugin I use to monitor the server on my phone can't connect

 

Attached are my diagnostics:

 

 

It's been difficult to diagnose, as the UI basically slows to a halt while it's having issues and I can't seem to find anything that pops out in the logs.

 

If anyone can take a look if I'm missing anything obvious in the logs, that would be greatly appreciated. I am not good at computer. Thanks in advance!

Edited January 19, 2021 by gregtamaki
remove attachments

[gregtamaki]

It appears that this is directly related to plex and sonarr with some sort of DB lockups when linux isos are downloading there.

 

From the Sonarr logs:

"System.Data.SQLite.SQLiteException (0x80004005): database is locked"

From the Plex logs:

"2021-01-19T18:00:17.988785522Z Sqlite3: Sleeping for 200ms to retry busy DB."

 

I set my downloaders to pause during normal plex watching hours and set plex to only scan for new episodes once every 12 hours. Seems to be fine so far - googling around for a real permanent solution seems like a futile effort. The general thinking is that Sqlite is a weakness in the plex armor and I should install emby or jellyfin. Leaving up for anyone else googling.