References
CPS-1856 - Getting issue details... STATUS
CPS-1857 - Getting issue details... STATUS
Out-of-scope
- Any other health check response other than Spring Boots(microservice) actuators is out of scope in this development.
Assumptions
# | Issue | Notes | Decisions |
|---|---|---|---|
| 1 | The service(name) of the DMI-plugin is being considered as if the name of microservice in the cluster. | The service(name) is being used during(within body part of) DMI registration process. | |
| 2 | The health check URL is the default Spring Boot actuator health check URL. | The format of the health check URL complies with Spring Boot actuator URL. | |
| 3 | The URL extension to service(name) is '/manage/health' | In the existing commit(development) '/manage/health' extension is being appended to service(name). | |
| 4 | The response body for that health check request is the generic response produced by Spring Boot health check actuator end-point. | The development relies on the default response created by Spring Boot health check actuator end-point. |
Issues & Decisions
# | Issue | Notes | Decisions |
|---|---|---|---|
| 1 | |||
| 2 |
Overview
The DMI Plugin Watchdog monitors the aliveness of DMI plugins. The job periodically (a configurable period of time) checks the health and status of each DMI plugin to ensure that
they are functioning properly. If a DMI-plugin is found to be unresponsive or in an unhealthy state, the Trust Level Cache will be updated with the latest status.
During the DMI-plugin registration process, DMI-plugin's service(name) will be added to Trust Level Cache with a value of COMPLETE. When the job runs, it will check aliveness of the all DMI-plugins
existing in the Trust Level Cache, by returning an UP or DOWN state for each service(name) in the response.
ncmp.timers.trust-evel.dmi-availability-watchdog-ms
Please note that the service(name) will be appended with a URL extension which is '/manage/health'.
service(name)/manage/health
A sample health check response generated by Spring Boot actuator health check end-point:
The 'status': 'UP' node is being considered in the development.