*: promote debugging metrics

[gyuho]

#9438

[xiang90]

can you explain why do we want to promote these metrics?

[gyuho]

@xiang90 Thought etcd_debugging_* has always been experimental? Mainly to make names consistent before improving our monitoring and alert guides.

Or should we make other metrics be consistent with etcd_debugging_*? Since we have etcd_server_version, defined right below etcd_debugging_server_lease_expired_total? We also have etcd_mvcc_hash_duration_seconds, which should be renamed to etcd_debugging_mvcc_hash_duration_seconds?

[gyuho]

ref. https://github.com/coreos/etcd/blob/master/Documentation/metrics.md#etcd_debugging-namespace-metrics

The metrics under the etcd_debugging prefix are for debugging. They are very implementation dependent and volatile. They might be changed or removed without any warning in new etcd releases. Some of the metrics might be moved to the etcd prefix when they become more stable.

We could serve etcd_debugging_* in parallel with etcd_* for smooth migration in v3.4, and deprecate debugging namespace in v3.5 (we do this for etcd --log-output flag).

[gyuho]

Will only promote etcd_debugging_mvcc_db_total_size_in_bytes in a backward compatible way (since it's used in https://github.com/coreos/etcd/blob/master/Documentation/op-guide/grafana.json#L344). I asked other production users, and found out quite a few are already using etcd_debugging_* metrics in production.