Kafka分区分布在一个消费者组中跨消费者
[英]Kafka partition distribution across consumers in a consumer group
问题描述
Am I right that, at every moment of time, partition distribution across consumers in a consumer group follows the Dirichlet box principle, if number of consumers is smaller than number of partitions?
我是对的，在任何时候，如果消费者数量小于分区数量，消费者组中消费者之间的分区分布遵循 Dirichlet 盒原则？ Ie if the consumer group has N consumers and a topic has M partitions and N < M, N-1 consumers got by one partition assigned to each, and the N-th consumer got M - N + 1 partitions.也就是说，如果消费者组有 N 个消费者，一个主题有 M 个分区，并且 N < M，则 N-1 个消费者通过分配给每个消费者的一个分区获得，而第 N 个消费者获得 M - N + 1 个分区。
1 个解决方案
解决方案1
0 2022-12-25 05:53:27
Some consumer will have multiple assigned partitions, yes.一些消费者会有多个分配的分区，是的。 It's not guaranteed to be the Nth-1 consumer that's started, or consecutive partitions assigned.不能保证是启动的第 N-1 个消费者，或分配的连续分区。
声明:本站的技术帖子网页，遵循CC BY-SA 4.0协议，如果您需要转载，请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.