This message was deleted.

https://apachedruidworkspace.slack.com/archives/C0303FDCZEZ/p1661287912383019?thread_ts=1661286488.092359&cid=C0303FDCZEZ mentions a couple of more configs you can check to speed up re-balancing and replication.

In addition to Rohan's suggestion, please make sure that you are using the correct strategy. If you heterogeneous historicals (i.e. historicals of different capacities), make sure to use

diskNormalized

. In other cases, you can stick with

cost

or

cachingCost

. https://druid.apache.org/docs/latest/configuration/index.html#coordinator-operation (see

druid.coordinator.balancer.strategy

)