在Centos 上安裝Kafka集群
安裝准備：
版本
Kafka版本：kafka_2.11-0.9.0.0
Zookeeper版本：zookeeper-3.4.7
Zookeeper 集群：bjrenrui0001 bjrenrui0002 bjrenrui0003
Zookeeper集群的搭建參見：在CentOS上安裝ZooKeeper集群
物理環境
安裝三台物理機：
192.168.100.200 bjrenrui0001（運行3個Broker）
192.168.100.201 bjrenrui0002（運行2個Broker）
192.168.100.202 bjrenrui0003（運行2個Broker）
該集群的創建主要分為三步，單節點單Broker,單節點多Broker,多節點多Broker
單節點單Broker
本節以bjrenrui0001上創建一個Broker為例
下載kafka:
下載路徑：http://kafka.apache.org/downloads.html
cd /mq/
wget http://mirrors.hust.edu.cn/apache/kafka/0.9.0.0/kafka_2.11-0.9.0.0.tgz
copyfiles.sh kafka_2.11-0.9.0.0.tgz bjyfnbserver /mq/
tar zxvf kafka_2.11-0.9.0.0.tgz -C /mq/
ln -s /mq/kafka_2.11-0.9.0.0 /mq/kafka
mkdir /mq/kafka/logs
配置
修改config/server.properties
vi /mq/kafka/config/server.properties
broker.id=1
listeners=PLAINTEXT://:9092
port=9092
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=1048576
socket.receive.buffer.bytes=1048576
socket.request.max.bytes=104857600
log.dirs=/mq/kafka/logs/kafka-logs
num.partitions=10
num.recovery.threads.per.data.dir=1
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
log.cleaner.enable=false
zookeeper.connect=bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181GG
zookeeper.connection.timeout.ms=6000
啟動Kafka服務:
cd /mq/kafka;sh bin/kafka-server-start.sh -daemon config/server.properties
或
sh /mq/kafka/bin/kafka-server-start.sh -daemon /mq/kafka/config/server.properties
netstat -ntlp|grep -E '2181|9092'
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
tcp6 0 0 :::9092 :::* LISTEN 26903/java
tcp6 0 0 :::2181 :::* LISTEN 24532/java
創建Topic:
sh /mq/kafka/bin/kafka-topics.sh --create --zookeeper bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181 --replication-factor 1 --partitions 1 --topic test
查看Topic：
sh /mq/kafka/bin/kafka-topics.sh --list --zookeeper bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181
producer發送消息:
$ sh /mq/kafka/bin/kafka-console-producer.sh --broker-list bjrenrui0001:9092 --topic test
first
message
consumer接收消息:
$ sh bin/kafka-console-consumer.sh --zookeeper bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181 --topic test --from-beginning
first
message
如果要最新的數據，可以不帶--from-beginning參數即可。
單節點多個Broker
將上個章節中的文件夾再復制兩份分別為kafka_2,kafka_3
cp -r /mq/kafka_2.11-0.9.0.0 /mq/kafka_2.11-0.9.0.0_2
cp -r /mq/kafka_2.11-0.9.0.0 /mq/kafka_2.11-0.9.0.0_3
ln -s /mq/kafka_2.11-0.9.0.0_2 /mq/kafka_2
ln -s /mq/kafka_2.11-0.9.0.0_3 /mq/kafka_3
分別修改kafka_2/config/server.properties以及kafka_3/config/server.properties 文件中的broker.id,以及port屬性，確保唯一性
vi /mq/kafka_2/config/server.properties
broker.id=2
listeners=PLAINTEXT://:9093
port=9093
host.name=bjrenrui0001
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=1048576
socket.receive.buffer.bytes=1048576
socket.request.max.bytes=104857600
log.dirs=/mq/kafka_2/logs/kafka-logs
num.partitions=10
num.recovery.threads.per.data.dir=1
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
log.cleaner.enable=false
zookeeper.connect=bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181
zookeeper.connection.timeout.ms=6000
vi /mq/kafka_3/config/server.properties
broker.id=3
listeners=PLAINTEXT://:9094
port=9094
host.name=bjrenrui0001
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=1048576
socket.receive.buffer.bytes=1048576
socket.request.max.bytes=104857600
log.dirs=/mq/kafka_3/logs/kafka-logs
num.partitions=10
num.recovery.threads.per.data.dir=1
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
log.cleaner.enable=false
zookeeper.connect=bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181
zookeeper.connection.timeout.ms=6000
啟動
啟動另外兩個Broker:
sh /mq/kafka_2/bin/kafka-server-start.sh -daemon /mq/kafka_2/config/server.properties
sh /mq/kafka_3/bin/kafka-server-start.sh -daemon /mq/kafka_3/config/server.properties
檢查端口:
[dreamjobs@bjrenrui0001 config]$ netstat -ntlp|grep -E '2181|909[2-9]'|sort -k3
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
tcp6 0 0 :::2181 :::* LISTEN 24532/java
tcp6 0 0 :::9092 :::* LISTEN 26903/java
tcp6 0 0 :::9093 :::* LISTEN 28672/java
tcp6 0 0 :::9094 :::* LISTEN 28734/java
創建一個replication factor為3的topic:
sh /mq/kafka/bin/kafka-topics.sh --create --zookeeper bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181 --replication-factor 3 --partitions 1 --topic my-replicated-topic
查看Topic的狀態:
$ sh /mq/kafka/bin/kafka-topics.sh --describe -zookeeper bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181 --topic my-replicated-topic
Topic:my-replicated-topic PartitionCount:1 ReplicationFactor:3 Configs:
Topic: my-replicated-topic Partition: 0 Leader: 3 Replicas: 3,1,2 Isr: 3,1,2
從上面的內容可以看出，該topic包含1個part,replicationfactor為3，且Node3 是leador
解釋如下：
"leader" is the node responsible for all reads and writes for the given partition. Each node will be the leader for a randomly selected portion of the partitions.
"replicas" is the list of nodes that replicate the log for this partition regardless of whether they are the leader or even if they are currently alive.
"isr" is the set of "in-sync" replicas. This is the subset of the replicas list that is currently alive and caught-up to the leader.
再來看一下之前創建的test topic, 從下圖可以看出沒有進行replication
$ sh /mq/kafka/bin/kafka-topics.sh --describe --zookeeper bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181 --topic test
Topic:test PartitionCount:1 ReplicationFactor:1 Configs:
Topic: test Partition: 0 Leader: 1 Replicas: 1 Isr: 1
多個節點的多個Broker
在bjrenrui0002、bjrenrui0003上分別把下載的文件解壓縮到kafka_4,kafka_5,kafka_6兩個文件夾中，再將bjrenrui0001上的server.properties配置文件拷貝到這三個文件夾中
vi /mq/kafka_4/config/server.properties
broker.id=4
listeners=PLAINTEXT://:9095
port=9095
host.name=bjrenrui0002
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=1048576
socket.receive.buffer.bytes=1048576
socket.request.max.bytes=104857600
log.dirs=/mq/kafka_4/logs/kafka-logs
num.partitions=10
num.recovery.threads.per.data.dir=1
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
log.cleaner.enable=false
zookeeper.connect=bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181
zookeeper.connection.timeout.ms=6000
vi /mq/kafka_5/config/server.properties
broker.id=5
listeners=PLAINTEXT://:9096
port=9096
host.name=bjrenrui0002
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=1048576
socket.receive.buffer.bytes=1048576
socket.request.max.bytes=104857600
log.dirs=/mq/kafka_5/logs/kafka-logs
num.partitions=10
num.recovery.threads.per.data.dir=1
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
log.cleaner.enable=false
zookeeper.connect=bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181
zookeeper.connection.timeout.ms=6000
vi /mq/kafka_6/config/server.properties
broker.id=6
listeners=PLAINTEXT://:9097
port=9097
host.name=bjrenrui0003
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=1048576
socket.receive.buffer.bytes=1048576
socket.request.max.bytes=104857600
log.dirs=/mq/kafka_6/logs/kafka-logs
num.partitions=10
num.recovery.threads.per.data.dir=1
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
log.cleaner.enable=false
zookeeper.connect=bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181
zookeeper.connection.timeout.ms=6000
vi /mq/kafka_7/config/server.properties
broker.id=7
listeners=PLAINTEXT://:9098
port=9098
host.name=bjrenrui0003
num.network.threads=3
num.io.threads=8
socket.send.buffer.bytes=1048576
socket.receive.buffer.bytes=1048576
socket.request.max.bytes=104857600
log.dirs=/mq/kafka_7/logs/kafka-logs
num.partitions=10
num.recovery.threads.per.data.dir=1
log.retention.hours=168
log.segment.bytes=1073741824
log.retention.check.interval.ms=300000
log.cleaner.enable=false
zookeeper.connect=bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181
zookeeper.connection.timeout.ms=6000
啟動服務
sh /mq/kafka/bin/kafka-server-start.sh -daemon /mq/kafka/config/server.properties
sh /mq/kafka_2/bin/kafka-server-start.sh -daemon /mq/kafka_2/config/server.properties
sh /mq/kafka_3/bin/kafka-server-start.sh -daemon /mq/kafka_3/config/server.properties
sh /mq/kafka_4/bin/kafka-server-start.sh -daemon /mq/kafka_4/config/server.properties
sh /mq/kafka_5/bin/kafka-server-start.sh -daemon /mq/kafka_5/config/server.properties
sh /mq/kafka_6/bin/kafka-server-start.sh -daemon /mq/kafka_6/config/server.properties
sh /mq/kafka_7/bin/kafka-server-start.sh -daemon /mq/kafka_7/config/server.properties
檢查：
$ netstat -ntlp|grep -E '2181|909[2-9]'|sort -k3
停服務：
sh /mq/kafka/bin/kafka-server-stop.sh
如果使用腳本停broker服務，會把單節點上的多broker服務都停掉，慎重！！！
ps ax | grep -i 'kafka\.Kafka' | grep java | grep -v grep | awk '{print $1}' | xargs kill -SIGTERM
到目前為止，三台物理機上的7個Broker已經啟動完畢：
[dreamjobs@bjrenrui0001 bin]$ netstat -ntlp|grep -E '2181|909[2-9]'|sort -k3
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
tcp6 0 0 :::2181 :::* LISTEN 24532/java
tcp6 0 0 :::9092 :::* LISTEN 33212/java
tcp6 0 0 :::9093 :::* LISTEN 32997/java
tcp6 0 0 :::9094 :::* LISTEN 33064/java
[dreamjobs@bjrenrui0002 config]$ netstat -ntlp|grep -E '2181|909[2-9]'|sort -k3
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
tcp6 0 0 :::2181 :::* LISTEN 6899/java
tcp6 0 0 :::9095 :::* LISTEN 33251/java
tcp6 0 0 :::9096 :::* LISTEN 33279/java
[dreamjobs@bjrenrui0003 config]$ netstat -ntlp|grep -E '2181|909[2-9]'|sort -k3
(Not all processes could be identified, non-owned process info
will not be shown, you would have to be root to see it all.)
tcp 0 0 0.0.0.0:2181 0.0.0.0:* LISTEN 14562/java
tcp 0 0 0.0.0.0:9097 0.0.0.0:* LISTEN 23246/java
tcp 0 0 0.0.0.0:9098 0.0.0.0:* LISTEN 23270/java
producer發送消息:
$ sh /mq/kafka/bin/kafka-console-producer.sh --broker-list bjrenrui0001:9092 --topic my-replicated-topic
consumer接收消息:
$ sh /mq/kafka_4/bin/kafka-console-consumer.sh --zookeeper bjrenrui0001:2181,bjrenrui0002:2181,bjrenrui0003:2181 --topic my-replicated-topic --from-beginning
