These are chat archives for atomix/atomix

11th
Aug 2018
Jordan Halterman
@kuujo
Aug 11 2018 00:45
#801 seems like a nice try
the primary-backup protocol may be a little more complicated
although maybe it can just use another primitive to manage sessions
Junbo Ruan
@aruanruan
Aug 11 2018 08:40
@kuujo thx,i found the annotation AtomixResource
Junbo Ruan
@aruanruan
Aug 11 2018 09:48
i run atomix cluster for serverval hours & using stress testing for times, cluster became unconected.
17:26:01.463 [raft-server-system-partition-2] ERROR i.a.p.raft.impl.RaftServiceManager - RaftServer{system-partition-2} - Cannot apply index 23466 (hasNext: true, nextInde: 1008)
17:26:01.463 [raft-server-system-partition-2] TRACE i.a.protocols.raft.roles.LeaderRole - RaftServer{system-partition-2}{role=LEADER} - Sending OpenSessionResponse{status=ERROR, error=RaftError{type=PROTOCOL_ERROR, message=Cannot apply index 23466}}
i add log for applyIndex of RaftServiceManager.java
it looks like the index roll over
reader next index became 1008, but current Index is bigger
Junbo Ruan
@aruanruan
Aug 11 2018 09:55
Wathing the cluster jm by VisualVm, the thread 'raft-server-system-partition-2' is running with messages
but i stopped all client
Johno Crawford
@johnou
Aug 11 2018 17:57
Yeah makes sense
Jordan Halterman
@kuujo
Aug 11 2018 19:02
I see this error every now and then but never with TRACE logs enabled, which I’d need to reproduce and fix it