Getting started

About five minutes, and the last step is the one that matters: a monitoring tool is worth nothing until you have watched it catch a failure.

1. Create a monitor

Sign in, open a project and add a monitor. It needs two things:

• A schedule — a cron expression, or a fixed interval. This is when Rippler expects to hear from the job, so it should match the job's real schedule.
• A grace duration — how late a ping may be before it counts as missed. Set it a little above the job's worst normal run time, so an unusually slow run does not page anyone. See schedules and grace.

The monitor starts as new: created, never pinged. It will not alert in that state.

2. Add the ping to your job

The monitor's page shows its ping URL and paste-ready snippets. The shortest thing that works is one call when the job finishes:

curl -fsS -m 10 --retry 5 -o /dev/null https://api.rippler.io/ping/<pingID>/

For a cron job, that goes in your own crontab, via crontab -e:

0 3 * * * /path/to/your-job.sh && curl -fsS -m 10 --retry 5 -o /dev/null "https://api.rippler.io/ping/<pingID>/"

Opening the run first buys you a duration and an immediate failure report, rather than waiting for the next ping to go missing:

curl … /ping/<pingID>/start/     # before the work
curl … /ping/<pingID>/           # it worked
curl … /ping/<pingID>/fail/      # it did not

The pings reference covers the URL shapes and their parameters.

3. Watch it report

Run the job by hand, or wait for its next scheduled run. The monitor goes to up, the ping count increases, and the run appears under Logs.

If nothing arrives, the job is very often fine and the ping is not — troubleshooting is a list of the usual reasons.

4. Break it on purpose

This is the step people skip. Send a failure by hand:

curl -fsS -m 10 -o /dev/null "https://api.rippler.io/ping/<pingID>/fail/"

The monitor goes down, and an alert goes to everyone on the team who has alert emails switched on. Now you know what an outage looks like before you have one, and that the address you expected it to reach actually receives it.

Send a success ping to clear it — a recovery alert follows the same way.

Then

• Have a lot of cron jobs already? rippler sync reads your crontab, creates a monitor for every job that does not have one, and rewrites those lines to report themselves.
• Noisy job that recovers by itself? Raise its failure tolerance instead of turning the monitor off.
• Working with a team? Alerts go to every member, and each person controls whether they receive them.