Details
Description
Steps to reproduce
- Created a cluster with 4 kv nodes, 2 search nodes and 1 query node
- Created 2 magma buckets with RAM Quota 1024MB each
- Started an internal XDCR replication between test_bucket and test_bucket2
- Started loading docs into the cluster
- Removed the query node and added the same node back with services index, query, cbas and backup
- Started rebalance - Rebalance succeeds
- But /pools/default/terseClusterInfo shows the status "isBalanced" as false
- Retried rebalance many times - the status for isBalanced never changes
Response for /pools/default/terseClusterInfo
{"clusterUUID": "03ee4155267399326af770c5dd3d831d","orchestrator": "ns_1@172.23.105.202","isBalanced": false,"clusterCompatVersion": "7.6"} |
/pools/default response mentions
servicesNeedRebalance": [{"code": "service_not_balanced","description": "Service needs rebalance.","services": ["fts"]}], |
A screen-recording for the same has been attached
This was not observed in regression runs or tests during 7.6.0-2090. Hence marking this as a regression
Attachments
Issue Links
- relates to
-
MB-60720 TLS: Rebalance 2 nodes during index build: service_rebalance_failed,fts agent_died, badmatch
- Closed