Maxing out CPU (from #535)

[apcrcom]

Posted about this recently and thought I had managed to sort it out but the problem is returned.

Cant run Hydra2 without CPU getting maxed out.

https://we.tl/t-X0G8rP7q1h

Tried setting memory to 1024.

Tried bypassing systemd for startup

Pretty much stumped at the moment. Would be grateful for any input.

Thanks

[theotherp]

Please set the log file level to debug and enable the logging marker "Performance". Restart. Check the system / bugreport tab to see what threads use CPU.

Make a screenshot and attach it with new debug infos. You can attach the files here.

[apcrcom]

Attached. Thankyou.

nzbhydra-debuginfos-2020-04-24-21-08.zip

[theotherp]

The database is writing a lot of data for some reason. I've had similar reports but haven't been able to reproduce this problem.

If you want you can play around a bit with the database settings:

1. Create a new folder config in the nzbhydra main folder and in there a file application.properties
2. Add this line spring.datasource.url=jdbc:h2:file:${nzbhydra.dataFolder:.}/database/nzbhydra;DEFRAG_ALWAYS=TRUE;WRITE_DELAY=20000;RETENTION_TIME=1000.
3. Try increasing the retention time. I think the default is 45 seconds, so 45000, perhaps choose 20000. Restart Hydra and see if that helps.

[apcrcom]

I did as you suggest and have run it couple days to see how it played out. Looks like its behaving now. Fingers xd.

Thanks for the support. Appreciate it.

[apcrcom]

Can I buy you a beer on Patreon or PayPal or something? (Sorry, I dont do bitcoin!)

…

On Fri, Apr 24, 2020 at 9:48 PM theotherp ***@***.***> wrote: The database is writing a lot of data for some reason. I've had similar reports but haven't been able to reproduce this problem. If you want you can play around a bit with the database settings: 1. Create a new folder config in the nzbhydra main folder and in there a file application.properties 2. Add this line spring.datasource.url=jdbc:h2:file:${nzbhydra.dataFolder:.}/database/nzbhydra;DEFRAG_ALWAYS=TRUE;WRITE_DELAY=20000;RETENTION_TIME=1000 . 3. Try increasing the retention time. I think the default is 45 seconds, so 45000, perhaps choose 20000. Restart Hydra and see if that helps. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#544 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGB6YVDST3ALCPXZCARTSPLROF7RJANCNFSM4MPX5RQA> .

[theotherp]

What values did you use? By the way, they can be configured in the GUI in the latest version (still in beta, though).

You can find donation options on the about page. Cheers.

[apcrcom]

I just used your suggestion

…

On Mon, Apr 27, 2020 at 2:28 PM theotherp ***@***.***> wrote: What values did you use? By the way, they can be configured in the GUI in the latest version (still in beta, though). You can find donation options on the about page. Cheers. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#544 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGB6YVG6JMIS2DQKH72MGF3ROUGI5ANCNFSM4MPX5RQA> .

[theotherp]

So you just changed the retention time from 1000 to 20_000? Cool...

[jnpeterson]

I've been having the same issue here - NZBHydra pegging the java process from 25-60% of the CPU consistently. I found this thread and set the logs as suggested and found my MVStore thread was pegged out as well. Followed the instructions given and changed the retention to 20000 and restarted. Now the java process is sitting between about 1-4%. Many thanks for this, it was slowing things down!