Jean-Daniel Cryans has posted comments on this change. Change subject: design-docs: multi-master for 1.0 release ......................................................................
Patch Set 3: (3 comments) Great work putting this all together. For even more completeness, it'd be nice to have a list of the in-memory caches and what feeds them. The information is scattered in the text at the moment. http://gerrit.cloudera.org:8080/#/c/2527/3/docs/design-docs/multi-master-1.0.md File docs/design-docs/multi-master-1.0.md: Line 183: (TBD) At least on the Java side, the code that detects which master is the leader was partially removed since it was buggy. Line 230: when a tserver detects that there's a new leader So they still need to "follow the leader"? Line 280: The tserver cache is expected to be warm very quickly as tservers are : now heartbeating to follower masters. Because they already know to heartbeat to it and don't have to figure out the new leader. In general, I dislike expecting things in big distributed systems to happen "very quickly". Since the followers are getting the heartbeats, could they keep some caches warm? Like the liveness one? Also, should ex-leaders clear their caches? -- To view, visit http://gerrit.cloudera.org:8080/2527 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: Iad76012977a45370b72a04d608371cecf90442ef Gerrit-PatchSet: 3 Gerrit-Project: kudu Gerrit-Branch: master Gerrit-Owner: Adar Dembo <[email protected]> Gerrit-Reviewer: Adar Dembo <[email protected]> Gerrit-Reviewer: Dan Burkert <[email protected]> Gerrit-Reviewer: David Ribeiro Alves <[email protected]> Gerrit-Reviewer: Jean-Daniel Cryans Gerrit-Reviewer: Kudu Jenkins Gerrit-Reviewer: Mike Percy <[email protected]> Gerrit-Reviewer: Todd Lipcon <[email protected]> Gerrit-HasComments: Yes
