Difference between revisions of "Monitoring (moved)"
Jump to navigation Jump to search
m (Mdaniel moved page Monitoring to Monitoring (moved): deprecation process in favor of sphinx and readthedocs)
Revision as of 16:37, 25 June 2012
linode.openhatch.orgis the main OpenHatch box, which runs the website.
linode2.openhatch.orgis the secondary server for OpenHatch. It hosts Nagios!
vm3.openhatch.orgis a third server, hosted at GPLHost, that runs the Jenkins continuous integration server.
- The Nagios configuration is owned by a user called nagios on linode2.openhatch.org.
- We use ssh keys for login.
- If you want SSH access to that account, file a bug requesting it, and attach an SSH key. You should hear back within 2 days; if you don't hear back by then, try to find paulproteus or jesstess on IRC.
- Then you can do:
- You'll know it's working if you are logged in. If you see a "Password:" prompt, then it is not working.
- Nagios notifications go to firstname.lastname@example.org. Anyone can subscribe to this list or read its archives.
In brief, here's what you need to know:
- Edit files in ~nagios/
- Once you know what changes you want to make, create a local branch with those changes:
git checkout -b my_changes
- As you make changes, make meaningful commits. Also, tell "git commit" to use your identity:
git commit --author="Some Body <email@example.com>"
- After you have made the changes, ask someone to review them and merge the changes to master.
- Rationale: If you stick to the above process, it is fairly easy to roll back to the "master" branch of the Nagios configuration.
- History: We came up with this process during issue332.
Viewing the web interface, and handling the daemon
~nagios/secrets/contains the mailman and Nagios web interface passwords.
- View the Nagios web interface at
- To restart the Nagios daemon, run
sudo /etc/init.d/nagios3 restart
In case of emergency
- See Emergency operations for the openhatch server. People with ssh keys set up for the Linode Shell (Lish) can reboot the box and have other limited emergency capabilities.
- Send Nagios notifications to IRC (
- Make the Nagios web interface world-viewable.
- Version the monitoring configurations.
- Send SMS alerts to people who want them.
- Add historical trending (Munin)?