Andrew Wong has posted comments on this change. Change subject: KUDU-2123. Auto-vivify cmeta on tombstoned replicas if doesn't exist at startup ......................................................................
Patch Set 7: (7 comments) http://gerrit.cloudera.org:8080/#/c/7988/7/src/kudu/consensus/consensus_meta-test.cc File src/kudu/consensus/consensus_meta-test.cc: PS7, Line 121: N nit: should be lowercased http://gerrit.cloudera.org:8080/#/c/7988/7/src/kudu/consensus/consensus_meta.h File src/kudu/consensus/consensus_meta.h: PS7, Line 175: is flushed to disk before returning This is no longer always true. PS7, Line 185: object from disk. IIUC, Create(NO_FLUSH_ON_CREATE) followed by Load() is an error, right? Is this by design / is it at all important? http://gerrit.cloudera.org:8080/#/c/7988/7/src/kudu/consensus/consensus_meta_manager-test.cc File src/kudu/consensus/consensus_meta_manager-test.cc: PS7, Line 103: Load Would using LoadOrCreate here as well be appropriate? http://gerrit.cloudera.org:8080/#/c/7988/7/src/kudu/consensus/consensus_meta_manager.h File src/kudu/consensus/consensus_meta_manager.h: PS7, Line 62: Returns an error if it cannot be found. Kind of follows from my other question: IIUC, Create(NO_FLUSH_ON_CREATE) followed by Load() is NOT an error here, right? If so, is this difference in APIs by design? It's a subtle difference that might be worth documenting. PS7, Line 71: ConsensusMetadataCreateMode create_mode = : ConsensusMetadataCreateMode::FLUSH_ON_CREATE, Just making sure I understand this, back-to-back calls to LoadOrCreate(NO_FLUSH_ON_CREATE) is not an error, right? It will just result in nothing on disk? http://gerrit.cloudera.org:8080/#/c/7988/7/src/kudu/tserver/ts_tablet_manager.cc File src/kudu/tserver/ts_tablet_manager.cc: PS7, Line 1088: guarantees that the replica has never voted in a leader election. I think I understand why this is true (voting --> flushing to disk). Does anything hinge on this assumption being true? -- To view, visit http://gerrit.cloudera.org:8080/7988 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I8ff6255b1fcbb12417b82853bcde9b239291492b Gerrit-PatchSet: 7 Gerrit-Project: kudu Gerrit-Branch: master Gerrit-Owner: Mike Percy <[email protected]> Gerrit-Reviewer: Adar Dembo <[email protected]> Gerrit-Reviewer: Alexey Serbin <[email protected]> Gerrit-Reviewer: Andrew Wong <[email protected]> Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Mike Percy <[email protected]> Gerrit-Reviewer: Tidy Bot Gerrit-Reviewer: Todd Lipcon <[email protected]> Gerrit-HasComments: Yes
