The scripts in this repository are wrappers for the OpsGenie API which will open and close OpsGenie incidents from the command line. These scripts are designed to help you integrate monit with OpsGenie.
This is based off of https://github.com/pinterest/pagerduty-monit
opsgenie-trigger <service-name>
opsgenie-trigger nginx on the server "webserver1.example.com"
will trigger a OpsGenie incident with the subject "nginx failed on
webserver1".
Requires opsgenie library for Python. You can run pip install opsgenie.
See https://docs.opsgenie.com/docs/opsgenie-python-api for more information.
To configure OpsGenie:
- login as an administrator
- Go to teams https://app.opsgenie.com/teams
- Select your team
- Select Integrations Tag
- Add Integration
- Select API
- Copy the API Key
- Replace the
YOUR_API_KEYinAPI_KEY = "YOUR_API_KEY"with your key in both scripts
Other Configuration parameters:
REALERT_AFTER_4_HOURS: By default, each alert will only be fired once. This allows the alert to re-fire after 4 hours.
To configure Monit to trigger an event with this script, save the script in a location like /etc/monit, make sure it's executable, and use an action like this in your monitrc file:
exec "/etc/monit/opsgenie-trigger <service-name>"
Example monit stanza to trigger an incident if nginx is not present:
check process nginx with pidfile /var/run/nginx.pid
if does not exist for 3 cycles
then exec "/etc/monit/opsgenie-trigger nginx"
else if passed for 3 cycles
then exec "/etc/monit/opsgenie-resolve nginx"
While the script is written for monit specifically, we use it for various alerts, especially for repeated triggers. The use is:
/etc/monit/opsgenie-trigger <summary> --description <description>
By default, the script is designed to fire new alerts only if one hasn't been recently fired. This is done via a file. You can disable this capability via --dontUseFiles.
Also, by default, the alert will not refire after 4 hours. You can enable this feature via --realert.