Details
-
Improvement
-
Resolution: Fixed
-
Critical
-
3.0, 4.0.0
-
Security Level: Public
-
Today couchbase clusters with 100 nodes have a chatty communication among nodes. This items tracks work needed to improve clusters of couchbase server over 100 nodes.
Description
Attachments
Issue Links
- relates to
-
MB-15573 DGM warmup gives different active/resident ratios per node per warmup (random)
- Resolved
-
MB-15990 KVShard::setLowPriorityVbSnapshotFlag not working as expected
- Resolved
-
MB-15757 Graceful failover either fails or takes very long time, delta rebalance fails - with latest build 3470
- Closed
-
MB-16146 cbcollect: Uncheck the selected node box
- Closed
-
MB-15303 overviewStats take 20 seconds to return on 6 node cluster with 10 buckets (was: Unable to view bucket status on "Cluster Overview" tab while running 10 buckets system test)
- Closed
-
MB-15676 Invalid read of size 4 in DCP Rollback task
- Closed
-
MB-16147 Consider change CPU util to be max CPU utilization
- Closed
-
MB-16227 Beam process not running - although service say cb is running
- Closed
-
MB-16447 Graceful failover failed in a highly scalable scenario of 128 data nodes
- Closed