Resuming rebalance after a failure
Improved rebalance latency of Couchbase Server is demonstrated through its ability to better resume rebalancing operation after any blips or failures have occurred in the process.
In case the rebalancing operation is interrupted at any point, it can be resumed instead of repeating the whole process.
Couchbase Server is able to perform incremental catch-up due to the restartability option provided through DCP (Database Change Protocol).
With DCP, the improved rebalance latency is achieved through efficient data movement, which is characterized by better capability to resume operation after failures, as well as to perform incremental catch-up.