Open
Description
A user reported the following issue (I checked the cluster state):
- He has around 50 Kafka sources. The biggest Kafka source has 4 desired pipelines, most of which have one.
- He has 10 indexers, each with 4-6 pipelines. 1 indexer has all the pipelines of the biggest Kafka source.
This is bad because, as you can imagine, only 1 indexer is responsible for indexing the biggest Kafka source. And all other indexers are doing less work.