Ok had the server crash on Saturday night - clients disconnected, webapp wouldn’t load, API calls failed from PRTG, etc. Wasn’t able to get in to do a diagnostics dump, but was able to grab the log files pre-reboot of the docker container. I’ve PM’d you those offline from this thread for security reasons. PRTG reported API timeouts at 11/17/2018 21:38:31 local time.
One of the things above you noted that I’d like to expand upon is the DB optimize. I’ve had several discussions about the SQLite DB over the past several months (example of one: SQLite Database Issues), as this error is extremely common throughout my logs.
My current configuration involves:
- the DB on a RAMDisk so it’s extremely fast and high IOPS. (miltiple GB/s throughput)
- I also optimize via the API call every 12 hours:
curl -XPUT http://10.1.1.4:32400/library/optimize - On top of that, I run a maintenance script every night when the docker container reboots: https://hastebin.com/ojayivafip.bash. If this needs modification or updating, I’m all ears as it was patched together via several forum/reddit posts, but my overall idea at this point is my DB is either too large or too complex for SQLite (2200 movies & 30,000 TV show files), and this is “just the way it is”.
Would love to know more about how I can fix my DB woes.
