As I remember, you can simply stop old broker, start the new one with the same broker id as the old one. It will start syncing replicas from other brokers and eventually will get all of them After this is done (all replicas are in sync) you can trigger leader election (or preferred replica election, whatever it is called) if it does not happen automatically.
---------- Andrey Yegorov On Thu, May 14, 2015 at 11:12 AM, Rajiv Kurian <ra...@signalfuse.com> wrote: > Hi all, > > Sometimes we need to replace a kafka broker because it turns out to be a > bad instance. What is the best way of doing this? > > We have been using the kafka-reassign-partitions.sh to migrate all topics > to the new list of brokers which is the (old list + the new instance - the > bad instance). Then we terminate the bad instance once we ensure that it is > getting no traffic. But it seems like this causes an unnecessary amount of > topic churn and is not equivalent to just moving the partitions the bad > broker was responsible for to the new instance. > > Is there a better way of going about replacing a single instance (not > adding capacity)? I'd ideally like to just be able to move the partitions > from the old broker to the new one instead of a complete rebalance. > > Thanks! >
