Details
Description
Per Alk, this change was intentional and made to fix a performance problem reported by Pavel (MB-11058) with just 2 replications causing "heavy CPU contention between memcached/ep-engine and XDCR replicators."
Creating this issue to vet performance of 32 vs 16 replicators per bucket and document the same if 16 is found to be ideal.
Attachments
Issue Links
- is triggered by
-
MB-11058 Failover during data load with enabled XDCR
- Closed