stantheman0128 commented on PR #10968: URL: https://github.com/apache/ozone/pull/10968#issuecomment-5280639171
@chihsuan thanks for running it against the mock site, and for the pointers. All three are addressed in `a954489`. The type reuse had one knock-on effect on `openKeysTable`, which I describe in the inline reply so it does not get lost in the diff. @yandrey321 good questions, thanks. On the `NA` row: the cluster level default and the bucket's own default are two separate things. `ozone.server.default.replication` and `ozone.server.default.replication.type` (`OMConfigKeys.java`) are applied when a key is written. The bucket's stored default is a different field, `optional DefaultReplicationConfig defaultReplicationConfig = 20` in the `BucketInfo` proto. `OMBucketCreateRequest` only reads it when `bucketInfo.hasDefaultReplicationConfig()` is true, and `OmBucketInfo` null guards it both when building the audit map and when building the protobuf. So a bucket created without an explicit replication setting stores nothing of its own, and the Recon bucket endpoint returns no `replicationConfigInfo` for it. `NA` is that case. It does not mean the bucket has no replication, it means the bucket carries no default of its own and writes fall back to the cluster setting. If you think `NA` reads as something stronger than that, I am happy to change the wording or leave the cell empty for those buckets. On EC-3-2 and EC-10-4: you are right that I had only exercised `rs-6-3-1024k`. I have added cases for `rs-3-2-1024k` and `rs-10-4-1024k`, plus one with a chunk size other than 1024k. The string is built from the serialized `codec`, `data`, `parity` and `ecChunkSize` fields, so the other supported combinations render the same way. -- 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]
