This page is a stub. Please consider expanding the article so it is complete.
This virtual machine hosts Ion.
The machine originally ran on a Gentoo distribution, but during a complete reinstall in June 2017 it was switched to Ubuntu.
|Hosted on||Hosted on::Antipodes|
|Provides service||Service provided::Intranet|
Production app info
App logs: /var/log/ion/
- Access: app_access.log
- Auth: app_auth.log
- Nginx: ion.tjhsst.edu.access.log
Database backups: /usr/local/db_backups/
- File format: YYYY-MM-DD-HHMM.pgdump.xz
- Hourly backups kept for the past month
- One backup a day for all previous months
App environment variables: /etc/supervisor/supervisord.conf
- DEBUG=FALSE, PRODUCTION=TRUE
App directory: /usr/local/www/intranet3/
- App settings: intranet/settings/
Set up environment:
$ ksu # workon ion # cd /usr/local/www/intranet3
Update production: (Changes should be tested in Ion-staging before pushing to production)
# fab deploy
Restart Ion worker processes:
# supervisord restart ion_async_workers:*
# /etc/init.d/nginx restart
Enable/disable maintenance mode:
# ./manage.py maintenance [on/off]
If for any reason the Ion virtual machine has been restarted, you will have to run the following commands to start Ion:
# mkdir -p /var/run/gunicorn/ # chown -R ion-data:ion-data /var/run/gunicorn/ # supervisorctl start ion_async # supervisorctl start ion_async_workers:* # supervisorctl status ion
To log everyone out of Ion, run the following command:
# redis-cli flushall
You will want to log everyone out of Ion after a restart, since their Kerberos caches (stored in /tmp) will be cleared, making them unable to access student information from LDAP until they are logged in again.
If you get a zsh error when running
# supervisorctl start ion_async_workers:*
this means zsh is trying to glob. To prevent this and allow supervisord to do the wildcard expansion, use noglob.
# noglob supervisorctl start ion_async_workers:*