Apache Ozone service health or status endpoints

[greenwich]

Is there any benefit of having the /health or /status endpoints for ozone services?

For example, would it be helpful for kubernetised environments so we can use that endpoint with a liveness or readiness probe?
Why do we need them? Because a real service health check might be more complicated than just pinging the HTTP/TCP port.

It might be helpful, for example, in the case of S3G.

It might be implemented as an MBean or a new servlet endpoint.
I have a simple implementation for S3G and can easily share it.

What do you think is the best way?

[ChenSammi]

Could you elaborate a bit more on the expected health/status info of S3G? S3G service is alive, data from S3G is readable, or data can be ingested into S3G?

[greenwich]

@ChenSammi, thanks for your response.

Service liveness might be a compound term; in my case, the health check should verify that the data is readable and can be ingested via S3G.

Probably, it means whether S3G can connect to OM and send and receive data.
In my mind, the easiest way to check that is to get the OzoneClient instance, get the object store and run a simple request to OM (like list volumes or something).