Michael Blow has posted comments on this change. Change subject: [NO ISSUE][HYR][*DB][CLUS] Startup lifecycle fixes ......................................................................
Patch Set 4: (4 comments) https://asterix-gerrit.ics.uci.edu/#/c/2004/4/asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/utils/ClusterStateManager.java File asterixdb/asterix-runtime/src/main/java/org/apache/asterix/runtime/utils/ClusterStateManager.java: PS4, Line 184: > who will notify waiters now? setState() already does notifyAll() PS4, Line 183: setState(ClusterState.UNUSABLE); > Let's add a log message after this saying that all cluster partitions are a Done PS4, Line 187: // if all storage partitions are active as well as the metadata node, then the cluster is active > fix this comment Done PS4, Line 190: setState(ClusterState.PENDING); > why do we do this? I assume to ensure we always transition through the PENDING state, otherwise we might go from UNUSABLE -> RECOVERING -> ACTIVE -- To view, visit https://asterix-gerrit.ics.uci.edu/2004 To unsubscribe, visit https://asterix-gerrit.ics.uci.edu/settings Gerrit-MessageType: comment Gerrit-Change-Id: Id30415325047008c013e305ca11ccbb76bc7d8d8 Gerrit-PatchSet: 4 Gerrit-Project: asterixdb Gerrit-Branch: master Gerrit-Owner: Michael Blow <[email protected]> Gerrit-Reviewer: Anon. E. Moose #1000171 Gerrit-Reviewer: Jenkins <[email protected]> Gerrit-Reviewer: Michael Blow <[email protected]> Gerrit-Reviewer: Murtadha Hubail <[email protected]> Gerrit-Reviewer: Till Westmann <[email protected]> Gerrit-HasComments: Yes
