devmadhuu commented on PR #9472: URL: https://github.com/apache/ozone/pull/9472#issuecomment-6489742940
> > It sounds great. Please do the follow-up. > > @devmadhuu , Actually, if you don't mind, we should redo it. We should avoid ambiguous terms like Healthy and Stuck. > > We could have serval dimensions like below. In each dimension, the states are mutually exclusive. > > * Replication: CORRECTLY_REPLICATED, UNDER_REPLICATED, OVER_REPLICATED, NO_ONLINE_REPLICA > > * Racks: INSUFFICIENT_RACKS, ENOUGH_RACKS > > * Open/close: OPEN_WITH_PIPELINE, OPEN_WITHOUT_PIPELINE, CLOSE, QUASI_CLOSED > > * Consistency: CONSISTENT, INCONSISTENT > > * Corruption: CORRUPTION_DETECTED, CHECKSUM_MATCHED > > * Recoverability: AUTO_RECOVERABLE, MANUAL_RECOVERY_REQUIRED > > > What do you think? Thanks @szetszwo for your proposal. On a broader context, it makes sense, but proposed states you mentioned and used in this PR were not created or redefined. During the development of ReplicationManager feature, some limited health check handlers for containers were being developed, so now they are not just a renaming exercise or defining the precise definitions, we need appropriate handlers in `ReplicationManager` with thought process on impacts, compatibility and existing QE and dev test cases. This needs a design discussion on broader area. Do you think we should create a separate tracker JIRA to revamp various health states being reported as part of RM as per your above proposal where we can have broader discussion on impact ? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
