2019独角兽企业重金招聘Python工程师标准>>> 
Kafka
1 简介
Kafka是一种高吞吐量的分布式发布订阅消息系统，它可以处理消费者规模的网站中的所有动作流数据。详细参考：apache kafka
2 快速启动
2.1 下载
https://www.apache.org/dyn/closer.cgi?path=/kafka/1.0.0/kafka_2.11-1.0.0.tgz
tar -xzf kafka_2.11-1.0.0.tgz
cd kafka_2.11-1.0.0/
2.2 启动zookeeper
参考：上一篇zookeeper集群启动
2.3 启动kafka server
2.3.1 修改config/server.properties
zookeeper.connect=192.168.10.11:2181,192.168.10.11:2182,192.168.10.11:2183
zookeeper 改成自己搭建的地址 zookeeper.connect=ip:port[,ip:port]
2.3.2 启动
bin/kafka-server-start.sh config/server.properties
默认启动端口：9092, 可以在配置文件中修改，启动过程可以看到kafka在zookeeper 中创建很多目录文件
[2017-12-15 11:35:20,632] INFO Creating /controller (is it secure? false) (kafka.utils.ZKCheckedEphemeral)
[2017-12-15 11:35:20,650] INFO Result of znode creation is: OK (kafka.utils.ZKCheckedEphemeral)
[2017-12-15 11:35:20,700] INFO [TransactionCoordinator id=0] Starting up. (kafka.coordinator.transaction.TransactionCoordinator)
[2017-12-15 11:35:20,702] INFO [TransactionCoordinator id=0] Startup complete. (kafka.coordinator.transaction.TransactionCoordinator)
[2017-12-15 11:35:20,703] INFO [Transaction Marker Channel Manager 0]: Starting (kafka.coordinator.transaction.TransactionMarkerChannelManager)
[2017-12-15 11:35:20,848] INFO Creating /brokers/ids/0 (is it secure? false) (kafka.utils.ZKCheckedEphemeral)
[2017-12-15 11:35:20,889] INFO Result of znode creation is: OK (kafka.utils.ZKCheckedEphemeral)
[2017-12-15 11:35:20,891] INFO Registered broker 0 at path /brokers/ids/0 with addresses: EndPoint(localhost,9092,ListenerName(PLAINTEXT),PLAINTEXT) (kafka.utils.ZkUtils)
在zookeeper 上查看
cd $ZOOKEEPER_DIR #自己zookeeper dir
./bin/zkCli.sh -server 192.168.10.11:2181 ##连接zookeeper client
查看目录
ls /
[cluster, controller, controller_epoch, brokers, admin, isr_change_notification, consumers, log_dir_event_notification, latest_producer_id_block, config]
2.4 创建一个topic
#创建一个topic = 'test'
bin/kafka-topics.sh --create --zookeeper 192.168.10.11:2181 --replication-factor 1 --partitions 1 --topic test
日志输出
[2017-12-15 11:48:03,335] INFO Created log for partition [test,0] in /tmp/kafka-logs with properties {compression.type -> producer, message.format.version -> 1.0-IV0, file.delete.delay.ms -> 60000, max.message.bytes -> 1000012, min.compaction.lag.ms -> 0, message.timestamp.type -> CreateTime, min.insync.replicas -> 1, segment.jitter.ms -> 0, preallocate -> false, min.cleanable.dirty.ratio -> 0.5, index.interval.bytes -> 4096, unclean.leader.election.enable -> false, retention.bytes -> -1, delete.retention.ms -> 86400000, cleanup.policy -> [delete], flush.ms -> 9223372036854775807, segment.ms -> 604800000, segment.bytes -> 1073741824, retention.ms -> 604800000, message.timestamp.difference.max.ms -> 9223372036854775807, segment.index.bytes -> 10485760, flush.messages -> 9223372036854775807}. (kafka.log.LogManager)
[2017-12-15 11:48:03,336] INFO [Partition test-0 broker=0] No checkpointed highwatermark is found for partition test-0 (kafka.cluster.Partition)
[2017-12-15 11:48:03,340] INFO Replica loaded for partition test-0 with initial high watermark 0 (kafka.cluster.Replica)
[2017-12-15 11:48:03,342] INFO [Partition test-0 broker=0] test-0 starts at Leader Epoch 0 from offset 0. Previous Leader Epoch was: -1 (kafka.cluster.Partition)
2.5 发送消息到topic
bin/kafka-console-producer.sh --broker-list localhost:9092 --topic test
2.6 启动消费者
bin/kafka-console-consumer.sh --bootstrap-server localhost:9092 --topic test --from-beginning
2.7 集群启动kafka
上面已经运行单个节点，但是生产环境最好还是运行一个集群，奇数个节点比较好。
从默认配置复制多两份配置出来
cp config/server.properties config/server-1.properties
cp config/server.properties config/server-2.properties
2.7.1 修改配置
config/server-1.properties:
broker.id=1 ##要求每个节点在集群中都是唯一的
listeners=PLAINTEXT://:9093 ##server 监听端口
log.dir=/tmp/kafka-logs-1 #日志数据目录
config/server-2.properties:
broker.id=2
listeners=PLAINTEXT://:9094
log.dir=/tmp/kafka-logs-2
2.7.2 启动这两个节点
bin/kafka-server-start.sh config/server-1.properties &
bin/kafka-server-start.sh config/server-2.properties &
2.7.3 对比单节点，集群中新建一个topic
bin/kafka-topics.sh --create --zookeeper 192.168.10.11:2181 --replication-factor 3 --partitions 1 --topic my-replicated-topic
2.7.4 查看topic描述信息
bin/kafka-topics.sh --describe --zookeeper 192.168.10.11:2181 --topic my-replicated-topic
Topic:my-replicated-topic PartitionCount:1 ReplicationFactor:3 Configs:
Topic: my-replicated-topic Partition: 0 Leader: 0 Replicas: 0,1,2 Isr: 0,1,2
leader : 这个节点负责所有的读写，再分配到不同的节点上
replicas : 当前还存活的节点broker.id
isr : in-sync 正在同步的节点
2.7.4.1 这时候查看单节点是新建的topic='test'
bin/kafka-topics.sh --describe --zookeeper 192.168.10.11:2181 --topic test
Topic:test PartitionCount:1 ReplicationFactor:1 Configs:
Topic: test Partition: 0 Leader: 0 Replicas: 0 Isr: 0
发现它只在我们brokerId=0上，并没有在集群中
2.7.4.2 启动消费生产者测试下
bin/kafka-console-producer.sh --broker-list localhost:9092 --topic my-replicated-topic
bin/kafka-consoconsumer.sh --bootstrap-server localhost:9092 --from-beginning --topic my-replicated-topic
和单节点比较结果并没有什么分别
2.7.4.2 模仿broker.id = 0 宕机会出现什么情况
netstat -nap | grep 9092
tcp6 0 0 :::9092 :::* LISTEN 11020/java
kill -9 11020
2.7.4.1 查看现在topic= my-replicated-topic的描述信息
bin/kafka-topics.sh --describe --zookeeper 192.168.10.11:2181 --topic my-replicated-topic
Topic:my-replicated-topic PartitionCount:1 ReplicationFactor:3 Configs:
Topic: my-replicated-topic Partition: 0 Leader: 1 Replicas: 0,1,2 Isr: 1,2
发现brokerId=0 还是在复制集set信息中
3 java kafka client 实例
3.1 maven 中引入依赖
org.apache.kafka
kafka-clients
1.0.0
org.apache.kafka
kafka-streams
1.0.0
3.2 producer 发送消息
public static void main(String[] args) throws Exception {
Properties props = new Properties();
//kafka servers
props.put("bootstrap.servers", "192.168.10.12:9092,192.168.10.12:9093,192.168.10.12:9094");
props.put("acks", "all");
props.put("retries", 0);
props.put("batch.size", 16384);
props.put("linger.ms", 1);
//topic 分组
props.put("client.id", "DemoProducer");
props.put("buffer.memory", 33554432);
//序列化工具
props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");
//序列化工具
props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");
Producer producer = new KafkaProducer<>(props);
for (int i = 0; i < 10; i++)
producer.send(new ProducerRecord("my-topic", Integer.toString(i), Integer.toString(i)));
producer.close();
}
}
配置信息参考：KafkaProducer javadocs
3.3 消费者
public static void main(String[] args) {
Properties props = new Properties();
//kafka servers
props.put("bootstrap.servers", "192.168.10.12:9092,192.168.10.12:9093,192.168.10.12:9094");
//group
props.put("group.id", "DemoConsumer");
props.put("enable.auto.commit", "true");
props.put("auto.commit.interval.ms", "1000");
props.put("key.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");
props.put("value.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");
KafkaConsumer consumer = new KafkaConsumer<>(props);
//订阅的topic
consumer.subscribe(Arrays.asList("my-topic"));
while (true) {
//超时时间 ms
ConsumerRecords records = consumer.poll(100);
for (ConsumerRecord record : records)
System.out.printf("测试 offset = %d, key = %s, value = %s%n", record.offset(), record.key(), record.value());
}
}
配置参数：KafkaConsumer javadocs