Sometimes mono flips and uses 50-100% CPU until jackett is stopped

[Redsandro]

Jackett version: v0.8.450
Mono version: 5.4.0.201
Docker image: docker-jackett

My computer was having issues. Upon inspection it turns out the temperature was near to 100 degrees Celsius. I've inspected sensor data and it started around the time I did a whole bunch of jackett searches.

I restarted jackett, and temperatures went back to normal. Then after some time mono flipped again. You can clearly see the temperature raise:

There were three mono processes running, two of them stuck racing between 50% and 100% CPU load. They stop and everything goes back to normal when Jackett is restarted.

I have no idea what causes this, and it doesn't happen frequently.

While I'm typing this and doing some searches in another tab, I see that this issue was reported here as well: linuxserver/docker-jackett#34

[kaso17]

If it happens again please provide the jackett log during the high CPU load time.
Is this a standard x86_64 platform or some embedded system?

[pboiseau]

I got the same things twice. I had to stop the container and start again to fix the issue.

[Redsandro]

@kaso17 in my case Intel Xeon Debian x86_64.

Any specific settings I need to enable to get a satisfactory log when this happens again?

[kaso17]

preferably with enhanced logging enabled (-t CLI option)

[supremacy2k]

My docker image seems to do the same.
And it seems to happen whenever a tracker crashes (going from working to failed status)
Can't provide logs as it's been a while since one of my trackers crashed.

But it least it's a place to start looking.

[pkoenig10]

I'm seeing the same thing on my server using the docker image.

My CPU usage is shown below. Red is Jackett and green is docker itself (ignoring the short spike around 02:15). Clearly nothing else is using CPU so the docker CPU usage is due to Jackett.

Here are the Jackett logs during startup and before shutdown.

[Quiksmage]

I had a similar issue and it was due to leaving auto-update on. I found similar reports on unraid's forums. Make sure that auto-update is turned off.

I'm running linuxserver's container.

If this can be fixed on jackett's side that'd be awesome, but I know, in general, you want to turn off updates within the container and wait for the container itself to be updated.

[kaso17]

@Quiksmage is right, if you're using docker and an update comes in it might cause these symptoms.

[pkoenig10]

Disable auto-update seemed to resolve the problem for me.

[kaso17]

Starting with version v0.8.676 jackett has a new command line option --NoUpdates which will prevent any kind of self updates.

[andycarver101]

Mono spiking in OSX, Checked disable auto update. Lets see who it goes...

[kaso17]

@andycarver101 another common cause of high CPU usage is the use of socks proxy, is that the case for you?

[deakster]

@kaso17 I have been noticing high CPU usage until restart since I configured the use of a socks proxy in Jackett. So yes, I think it is related.

[xboxhacks]

Same here, SOCKS 5 proxy has made my cpu usage hit over 70% on my synology nas