Apache Zookeeper is a centralized service for maintaining configuration information providing distributed synchronization. ZooKeeper is used in Kafka for cluster management and to maintain the topics information synchronized.
Adjust the following parameters:
zookeeper.connect=127.0.0.1:2181,192.168.0.32:2181 zookeeper.session.timeout.ms=6000 zookeeper.connection.timeout.ms=6000 zookeeper.sync.time.ms=2000
Here is an explanation of these settings:
zookeeper.connect: Default value: null. This is a comma-separated value in the form of thehostname:portstring, indicating the Zookeeper connection. Specifying several connections ensures the Kafka cluster reliability and continuity. When one node fails, Zookeeper uses the chroot path (/chroot/path) to make the data available under that particular path. This enables having the Zookeeper cluster available for multiple Kafka clusters. This path must be created before starting the Kafka cluster, and consumers must use the same string.zookeeper.session.timeout.ms: Default value: 6000. Session timeout means that if in this time period a heartbeat from the server is not received, it is considered dead. This parameter is fundamental, since if it is long and if the server is dead the whole system will experience problems. If it is small, a living server could be considered dead.zookeeper.connection.timeout.ms: Default value: 6000. This is the maximum time that the client will wait while establishing a connection to Zookeeper.zookeeper.sync.time.ms: Default value: 2000. This is the time a Zookeeper follower can be behind its Zookeeper leader.
- From the Kafka perspective, the ZooKeeper parameters are detailed at: http://kafka.apache.org/documentation.html#brokerconfigs
- The Apache ZooKeeper home page is here: https://zookeeper.apache.org/