Increase gunicorn's worker timeout since some /admin commands take a long time

This commit is contained in:
Joshua Tauberer 2022-10-08 08:17:24 -04:00
parent 22a6270657
commit 89cd9fb611

View file

@ -78,6 +78,8 @@ rm -f /tmp/bootstrap.zip
# Create an init script to start the management daemon and keep it
# running after a reboot.
# Set a long timeout since some commands take a while to run, matching
# the timeout we set for PHP (fastcgi_read_timeout in the nginx confs).
# Note: Authentication currently breaks with more than 1 gunicorn worker.
cat > $inst_dir/start <<EOF;
#!/bin/bash
@ -93,7 +95,7 @@ chmod 640 /var/lib/mailinabox/api.key
source $venv/bin/activate
export PYTHONPATH=$(pwd)/management
exec gunicorn -b localhost:10222 -w 1 wsgi:app
exec gunicorn -b localhost:10222 -w 1 --timeout 630 wsgi:app
EOF
chmod +x $inst_dir/start
cp --remove-destination conf/mailinabox.service /lib/systemd/system/mailinabox.service # target was previously a symlink so remove it first