junjizhi/redash-datadog

A datadog plug-in for redash

redash-datadog

A datadog plug-in for redash. The plug-in has three parts:

• A datadog agent check that collects redash-related metrics from redash HTTP status endpoint, with metrics including dashboard counts, queries count, query result count, etc.

• Celery and gunicorn process metrics. It may be useful to monitor the current number of these worker processes.

• Celery and gunicorn process age. The check ProceeAgeCheck.py will send the oldest process age to datadog. It may be useful to know how long these processes have been running.

Usage

For redash status agent check,

• Copy redash.py to /etc/dd-agent/checks.d/
• Put in the admin api key to redash.yaml.example and copy it to /etc/dd-agent/conf.d/redash.yaml.
• Restart dd-agent

For process checks, tune the parameters and copy process.yaml to /etc/dd-agent/conf.d/. You can find documentation here. Then restart dd-agent. Now you can add graphs or monitors in the datadog web console. For example, you can search the metric system.processes.threads with tag celery.

For process age checks, copy ProcessAgeCheck.py and check_proc_age.sh to /etc/dd-agent/checks.d/ and yaml to the conf.d/ path. Restart dd-agent.

Credits

• Thanks for Kevin Martin (Melraidin) for the idea and the code review.

References

• http://docs.datadoghq.com/guides/agent_checks/
• http://docs.redash.io/en/latest/usage/maintenance.html#monitoring
• https://github.com/DataDog/dd-agent/blob/master/conf.d/process.yaml.example