WebUI becomes unreachable

[shiftylilbastrd]

Not sure when exactly it started but for some reason the WebUI has become unreachable. I'm running the latest build in docker. When I start the container everything seems to be working fine but at some point that I have yet to be able to pin down it becomes unreachable and the container shows as unhealthy. Tried looking in the logs but there are so many different ones and some that seem super old.

[shiftylilbastrd]

Traced this down to Whisparr v3 as the issue. I Disabled all the services and reenabled them one at a time until I recreated the problem. Eros.log is over 2.2GB. Neither Notepad or VSCode will open it because it's so large.

[shiftylilbastrd]

I have logging level set to warning and the logs are full of Debug messages

[enniobotti]

I have the very same issue...

[Admin9705]

it's whisparr v3. next focus - disable for now. will implement a cool off period to assist. will have to deploy whisparr v3 again and see what happens.

[Admin9705]

@enniobotti @shiftylilbastrd - does huntarr 9.1+ fix it? I told it to stop spamming logs and etc, so... fingers crossed?

[Arturoe1]

now it takes forever to load any settings, it seems it has new memory leaks too.

[Scags104]

i had a similar issue to the point my server was unreachable locally even via ping however some of my other dockers which were hosted externally still loaded and worked properly. Not a scenario i thought was even possible. I wiped my server and started from scratch.... install Huntarr... same issues within a few minutes of deploying.

Turns out i had to change/expand my docker ip block. Example
sudo nano /etc/docker/daemon.json

add:

{
  "default-address-pools": [
    {
      "base": "10.xxx.0.0/16", 
      "size": 24
    }
  ]
}

note: the 10.x.x.x block above should be safe just make sure it doesnt interfere with what you have going on locally.

save, restart docker daemon service and then bring down and back up all containers to grab the new ip block.
Huntarr has been working fantastic since then.

[enniobotti]

Sorry for the late reply! I only managed to have a look at it now. I am running v.9.1.9.1. The issue is not fixed. Log is still running haywire...

[JadElClemens]

Having this same issue running on Arch from AUR (not Docker), I have only configured Sonarr and Radarr and get a totally unresponsive web server after just logging in and loading the homepage. After that I'm not able to navigate anywhere else before it locks up.

[JadElClemens]

Confirmed that this seems to be a logging issue, I made a few manual changes in the Python source on my machine and the web UI is much more stable.

On Arch package from AUR, modified /usr/share/webapps/huntarr/src/primary/utils/logger.py:

• Hacked DebugLogsFilter to always return False (just a shot in the dark)
• Replace all log levels under the comment Always use DEBUG level[...] from logging.DEBUG -> logging.WARNING (probably an overly aggressive change). There were three instances.

Huntarr still crashes from time to time, but it's due to some db error. The Web UI is much much much more responsive after this change.