Details
Description
Step to reproduce:
- make some viewChanges
- demote node1 with current viewNo X
- Simultaneously stop indy-node on all nodes (exclude demoted node)
- Simultaneously start indy-node on all nodes (exclude demoted node)
- Promote node1
Actual results:
node1 cannot link with any viewChange from pool, until it reach #(X+1) (more than current viewChange on node1).
Expected results:
Re-promoted node accept viewNo from other nodes
2018-02-27 13:19:33,584 | INFO | message_processor.py (29) | discard | Node1 discarding message INSTANCE_CHANGE{'reason': 26, 'viewNo': 1} because Received instance change request with view no 1 which is not more than its view no 4
Acceptance Criteria:
Create a plan of attack, and raise appropriate stories and epics that can be scheduled.
Attachments
Issue Links
- is duplicated by
-
INDY-1198 Pool stopped working if primary node was not included to schedule in upgrade transaction
-
- Complete
-
- relates to
-
INDY-1256 STN lost consensus
-
- Complete
-
-
INDY-1198 Pool stopped working if primary node was not included to schedule in upgrade transaction
-
- Complete
-
-
INDY-1422 Part of nodes continued ordering txns after `incorrect state trie` under load
-
- Complete
-
-
INDY-1465 Concurrent view change
-
- Complete
-
-
INDY-794 Primary selection possibly failure cases
-
- Complete
-