Where does zookeeper and Kafka fit in hadoop 2.6 cluster -
hadoop 2.6 uses yarn next generation map reduce , cluster manager. still need use zookeeper hadoop 2.6 cluster managing services? how setup zookeeper.
how kafka connectivity installed hadoop cluster. consumer , producer kafka send data hadoop file system.
where fit in.
i have setup hadoop 2.6 single node cluster. next, way understand have zookeeper , kafka data streaming hadoop file system. , don't have idea how use kafka hadoop or api.
zookeeper coordination framework distributed systems. zookeeper used coordinating state in hdfs & yarn high availability, coordination between hbase master , region servers etc. kafka works in combination apache storm, apache hbase , apache spark real-time analysis , rendering of streaming data. common use cases include:
- stream processing.
- website activity tracking
- metrics collection , monitoring
- log aggregation
usually use kafka along storm. storm needs zookeeper cluster coordination between nimbus , supervisor. kafka need zookeeper storing information cluster status , consumer offsets.
basically zookeeper provides highly available file system users/application can read/write small data. data can related communication or transactions. since file system highly available, communications complete , not go partial or unknown state. zookeeper cluster can withstand upto number of failures depending upon number of partitions(say n), can tolerate n-1 failures. more details, can refer following urls 1 2 3
Comments
Post a Comment