Through the kafka manager, you can see that for the partition with partition 2, its Preferred Leaderd status has become true. triggered> done, in the following format- Can also be usedkafka managerto view thetopicofpartitiondistribution, we can seepartitionas2ofthe partition leadernot a priority copy of a copy. We expect to re-load balance the topic leader and let the partition choose "preferred replica" as the leader. When creating a topic, Kafka tries to divide the partition equally among all brokers, and divide the replicas equally among different brokers. kafka kubernetes manager devops done right 
You'll know when you see it, Congratulations on the launch of the Sought Tech site. kafka Kafka broker id 200 2 broker.id listeners Kafka 1) kaf 1) kafkakafka Partition Topic Partition Broker Partition Repl Kafka Kafka KafkaLinkedinpartitionreplicazookeeper 118 West 21th Street, Suite 980 New York,NY 10110, Copyright 2021; All rights reserved | www.soughttech.com, java.net.NoRouteToHostException in zookeeper: No route to host, How to ensure that RocketMQ messages are not lost, In-depth analysis of the differences between Linux and Windows systems, What are the advantages of iOS development? Kafka Producer web Page View CPUMemo Kafka LinkedIn Apache Kafka, Kafka https://www.cnblogs.com/tree1123/p/11150927.html, Kafka Kafka Kafka . The format of xx.json is explained as follows: partitions for which preferred partitions for which preferred
--path-to-json-file
{"partitions":[{"topic":"topic_replica_test","partition":0},{"topic":"topic_replica_test","partition":1},{"topic":"topic_replica_test","partition" :2}]}, Use theelection.jsonfile to perform priority copy elections, /opt/cloudera/parcels/KAFKA/lib/kafka/bin/kafka-preferred-replica-election.sh --zookeeper zk1:2181 --path-to-json-file election.json. All Rights Reserved. 
Moreover, Kafka allows partitions to maintain a certain imbalance.The priority copy balance of a single topic does not represent the priority copy balance of the entire cluster. The kafka-preferred-replica-election command is used to rebalance the Leader load, Usage 1: Trigger load balancing for all topic Leaders, Usage 2: Trigger load balancing for a topic Leader, Automatic load balancing of topic Leader can be realized by configuring auto.leader.rebalance.enable=true, Reference : https://blog.csdn.net/lingbo229/article/details/83375066, kafka-preferred-replica-election --zookeeper cdh-002/kafka, kafka-preferred-replica-election --zookeeper cdh-002/kafka --path-to-json-file xx.json replica leader election needs to be replica leader election should be {"topic": "foobar", "partition": 2}] For manual priority copy elections, it is also recommended to conduct the election in batches to avoid simultaneous priority copy elections for multiple topics with large data volumes. 
In production, it is not recommended to set this parameter to true. The leader replica is responsible for all reads and writes. First, checkthepartitiondistribution oftopic, /opt/cloudera/parcels/KAFKA/lib/kafka/bin/kafka-topics.sh --zookeeper zk1:2181 --topic topic_replica_test --describe. However, over time, the broker may be down, which will lead to leader migration, resulting in unbalanced load on the cluster.
Option Description
Provide optimal solutions to technical problems. [root@cdh-003 ~]# kafka-preferred-replica-election Check againtopicofpartitiondistribution, discoverypartitionfor the2-pointarea leaderhas been adjusted to151for theleaderof. }. {"partitions":
Copyright 2013 - 2022 Tencent Cloud. [{"topic": "foo", "partition": 1}, All replicas of each partitiion are called "assigned replicas", the first replica in "assigned replicas" is called "preferred replica", and the newly created topic is generally "preferred replica" as leader. Because of the automatic leader replica election, in a large data volume environment, it may cause client congestion.