Rebalance partitions on MapR Streams

[Vinayak_Meghraj]

Is there some time factor involved in partitions being reallocated during consumer crash or consumer pid killed. How long does it take to rebalance partitions on MapR Streams and is there any configuration property to control the behaviour ?

[Vinayak_Meghraj · Accepted solution]

Whenever any consumer is added or existing consumer crashed , rebalancing of partitions happens.

This is the largest window allowed for the rebalancing phase, where clients are not reading anything from Kafka.
rebalance.backoff.ms * rebalance.max.retries.

But both the above properties are ignored on MapR Streams and instead streams.rpc.timeout.ms is used.

streams.rpc.timeout.ms - Specifies the length of time in milliseconds to wait for a response from the HPE Ezmeral Data Fabric Event Store server.

Differences between HPE Ezmeral Data Fabric Event Store and Apache Kafka Configuration has been documented :

https://docs.datafabric.hpe.com/62/MapR_Streams/differences_in_configuration_parameters_for_producers_and_consumers.html?hl=streams.rpc.timeout.ms