mirror of
https://gitlab.com/allianceauth/allianceauth.git
synced 2025-07-11 13:30:17 +02:00
51 lines
2.5 KiB
Markdown
51 lines
2.5 KiB
Markdown
# Troubleshooting
|
|
|
|
## Something broken? Stuck on an issue? Can't get it set up?
|
|
|
|
Start here:
|
|
- read the [documentation](https://github.com/R4stl1n/allianceauth/wiki)
|
|
- check the [issues](https://github.com/R4stl1n/allianceauth/issues?utf8=%E2%9C%93&q=is%3Aissue) - especially closed ones
|
|
- check the [forums](https://forums.eveonline.com/default.aspx?g=posts&t=383030)
|
|
|
|
No answer?
|
|
- open an [issue](https://github.com/R4stl1n/allianceauth/issues)
|
|
- harass us on [gitter](https://gitter.im/R4stl1n/allianceauth)
|
|
- post to the [forums](https://forums.eveonline.com/default.aspx?g=posts&t=383030)
|
|
|
|
## Common Problems
|
|
|
|
### `pip install -r requirements.txt` is failing
|
|
|
|
Most commonly, your repositories did not include the `requests` package. Install it and try again: `sudo pip install requests`
|
|
|
|
Otherwise it's usually a missing dependency. Check [the list](../installation/auth/dependencies.md), reinstall, and try again.
|
|
|
|
### I'm getting an error 500 trying to connect to the website on a new install
|
|
|
|
Read the apache error log: `sudo nano /var/log/apache2/error.log`
|
|
|
|
If it talks about failing to import something, google its name and install it.
|
|
|
|
If it whines about being unable to configure logger, make sure the log directory is write-able: `chmod -R 777 /home/allianceserver/allianceauth/log`, then reload apache.
|
|
|
|
### Groups aren't syncing to services
|
|
|
|
Make sure the background processes are running: `ps aux | grep celery` should return more than 1 line. More lines if you have more cores on your server's processor. If there are more than two lines starting with `SCREEN`, kill all of them with `kill #` where `#` is the process ID (second column), then restart with [these background process commands](../installation/auth/quickstart.md) from the allianceauth directory. You can't start these commands as root.
|
|
|
|
If that doesn't do it, try clearing the worker queue. First kill all celery processes as described above, then do the following:
|
|
|
|
sudo rabbitmqctl stop_app
|
|
sudo rabbitmqctl reset
|
|
sudo rabbitmqctl start_app
|
|
python manage.py celeryd --purge
|
|
|
|
Press control+C once.
|
|
|
|
python manage.py celeryd --discard
|
|
|
|
Press control+C once.
|
|
|
|
Now start celery again with [these background process commands.](../installation/auth/quickstart.md)
|
|
|
|
While debugging, it is useful to see if tasks are being executed. The easiest tool is [flower](http://flower.readthedocs.io/). Install it with this: `sudo pip install flower`, then start it with this: `celery flower --broker=amqp://guest:guest@localhost:5672//`. To view the status, navigate to your server IP, port 5555.
|