High disk I/O usage

[Daniel15]

NZBHydra is consistently reading/writing at ~60MB/s even when it's just sitting idle.

System load dropped quite a bit once I stopped my NZBHydra Docker container.

• Docker container: lscr.io/linuxserver/nzbhydra2, latest version as of today
• OS: Unraid 6.12.4
• debuginfos: nzbhydra-debuginfos-2024-02-18-18-20.zip

[theotherp]

Yeah, that's insane. I know of some cases where the DB causes high IO but I don't think I've ever seen it this high.

Thanks for the debug infos but I need some more data. Enable the display of advanced options in the config and then enable the logging marker for performance:

Then when the IO usage is high go to the debug infos section and make a screenshot. It should look a bit like this: , only "hopefully" with more activity. That may allow me to see if it's the database thread doing the work.

Then create a thread dump by running this curl: curl 'http://127.0.0.1:5076/actuator/threaddump' -i -X GET -H 'Accept: text/plain' > threaddump.txt (using whatever host and IP is correct) and post the resulting txt file.

Edit: I just remembered the thread dump will probably return an empty result. See if it works.

[Daniel15]

I had to reinstall all my Docker containers and now it's not doing it any more. I see zero I/O from NZBHydra. I'll keep an eye out for it and follow your steps if I see it again. Maybe it only happens after running for a while.