kafka命令如下:
kafka-shell基本命令
在节点hadoop-2,hadoop-3,hadoop-5,启动kafka
启动命令如下
kafka-server-start.sh /usr/local/kafka_2.11-0.10.0.1/config/server.properties > /usr/local/kafka_2.11-0.10.0.1/logs/logs &
1.创建topic
kafka-topics.sh --zookeeper hadoop-2:2181,hadoop-3:2181,hadoop-5:2181 --create --topic your.topic.name --partitions 30 --replication-factor 1
partitions指定topic分区数，replication-factor指定topic每个分区的副本数（一般等于broker个数）
2.查看topic列表
kafka-topics.sh --zookeeper hadoop-2:2181,hadoop-3:2181,hadoop-5:2181 --list
3.查看topic信息
kafka-topics.sh --zookeeper hadoop-2:2181,hadoop-3:2181,hadoop-5:2181 --describe --topic your.topic.name
4.向topic生产数据
kafka-console-producer.sh --broker-list hadoop-2:9092,hadoop-4:9092,hadoop-5:9092 --topic your.topic.name
5.消费者消费数据
kafka-console-consumer.sh --zookeeper hadoop-2:2181,hadoop-3:2181,hadoop-5:2181 -topic your.topic.name --from-beginning
6.查看topic某分区偏移量的最大（小）值
kafka-run-class.sh kafka.tools.GetOffsetShell --topic kafkademo --time -1 --broker-list hadoop-2:9092,hadoop-4:9092,hadoop-5:9092 --partitions 0
7.增加topic分区数量（只能增加，无法减少）
kafka-topics.sh --zookeeper hadoop-2:2181,hadoop-3:2181,hadoop-5:2181 --alter --topic your.topic.name --partitions 40
8.查看kafka消费进度
kafka-run-class.sh kafka.tools.ConsumerOffsetChecker --zookeeper hadoop-2:2181,hadoop-3:2181,hadoop-5:2181 --group pv
9.删除kafka的队列[注意需要重启kafka集群]
kafka-run-class.sh kafka.admin.DeleteTopicCommand --topic test_kafka --zookeeper chenx02:2181
10.查看不可用的分区
kafka-topics.sh --describe --unavailable-partitions --zookeeper chenx02:2181 --topic test_kafka
11.发送消息
./kafka-console-producer.sh --broker-list chenx02:9092 --topic test
12.---查看kafka的数据偏移量 [root@hadoop-5 data]#
kafka-run-class.sh kafka.tools.GetOffsetShell --topic guaishou --time -1 --broker-list 192.***:9092 --partitions 0
2.hadoop和zookeeper脚本管理集群
starthadoop.sh脚本
#!/bin/bash
ssh hadoop-2 "/usr/local/zookeeper/zookeeper-3.4.8/bin/zkServer.sh start"
sleep 5s
ssh hadoop-3 "/usr/local/zookeeper/zookeeper-3.4.8/bin/zkServer.sh start"
sleep 5s
ssh hadoop-5 "/usr/local/zookeeper/zookeeper-3.4.8/bin/zkServer.sh start"
sleep 5s
/usr/local/hadoop/hadoop-2.7.3/sbin/start-dfs.sh
sleep 30s
ssh hadoop-3 "/usr/local/hadoop/hadoop-2.7.3/sbin/start-yarn.sh"
sleep 30s
ssh hadoop-2 "/usr/local/spark/spark-2.2.1-bin-hadoop2.7/sbin/start-all.sh"
stophadoop.sh
#!/bin/bash
ssh hadoop-2 "/usr/local/spark/spark-2.2.1-bin-hadoop2.7/sbin/stop-all.sh"
sleep 10s
ssh hadoop-3 "/usr/local/hadoop/hadoop-2.7.3/sbin/stop-yarn.sh"
sleep 30s
/usr/local/hadoop/hadoop-2.7.3/sbin/stop-dfs.sh
sleep 30s
ssh hadoop-5 "/usr/local/zookeeper/zookeeper-3.4.8/bin/zkServer.sh stop"
sleep 3s
ssh hadoop-3 "/usr/local/zookeeper/zookeeper-3.4.8/bin/zkServer.sh stop"
sleep 3s
ssh hadoop-2 "/usr/local/zookeeper/zookeeper-3.4.8/bin/zkServer.sh stop"
sleep 3s
------------------------------------------------------------------------------------------------------------------------------------------
1)关于模拟kafka消费的博客（spring+springMVC+mybatis+kafka）。http://www.cnblogs.com/jun1019/p/6580371.html