Details
-
Bug
-
Resolution: Fixed
-
Critical
-
7.2.5, 7.6.2
-
Untriaged
-
0
-
Unknown
-
Analytics Sprint 47
Description
Discovered during testing of Goldfish features, when the CC node is rebalanced out, it is always the topology change orchestrator selected by ns_server. As a result, once the rebalance out is complete there are no nodes responsible for singleton tasks managed by the 'cbas' process, including monitoring of server groups updates & recording IP/version changes by the nodes.
This is already fixed in Goldfish, but is still broken in 7.2.6/7.6.3 at present.
Attachments
Issue Links
- is duplicated by
-
MB-62745 [Analytics][K8S] : Service 'cbas' exited with status 2
- Resolved
For Gerrit Dashboard: MB-62845 | ||||||
---|---|---|---|---|---|---|
# | Subject | Branch | Project | Status | CR | V |
213320,2 | MB-62845: select highest priority node after orchestrator rebalanced out | neo | cbas | Status: MERGED | +2 | +1 |
213335,1 | MB-62845,MB-62744: merge branch 'neo' into 'trinity' | trinity | cbas | Status: MERGED | +2 | +1 |
213336,1 | MB-62845,MB-62744: merge branch 'neo' into 'trinity' | trinity | cbas-core | Status: MERGED | +2 | +1 |