Enable health monitoring of Ceilometer
Ceilometer can be a key service whose health needs to be monitored. This blueprint will describe a pluggable entry point that will be instrumented at critical points to reflect the health of the Telemetry (Ceilometer) service itself.
Blueprint information
- Status:
- Complete
- Approver:
- None
- Priority:
- Undefined
- Drafter:
- Rob Raymond
- Direction:
- Needs approval
- Assignee:
- Rob Raymond
- Definition:
- Obsolete
- Series goal:
- None
- Implementation:
- Unknown
- Milestone target:
- None
- Started by
- Completed by
- gordon chung
Related branches
Related bugs
Sprints
Whiteboard
Use Cases:
The main use case for this blueprint is to support load balancer health checks. The load balancer will ping Ceilometer to make sure the service is available, otherwise it will take it off from the list of balanced nodes. Ceilometer will implement a "public" endpoint interface that returns a 200 OK in case the system has good health or a 500 detailing the issue. The health check endpoint should validate that all the components within Ceilometer are working correctly before reporting a 200. Malfunctioning, or lack of connectivity with some of the subsystems should trig an error response since the system may be unstable at the time of the call.
closing, we don't have an api. jd has some magic proposal in his head to apparently monitor the agents we have. we can track that as wishlist if he ever does it. - gordc(2017-12)