Centos7.5安装kafka集群

Tags： kafka

Centos7.5安装kafka集群

主机环境

基本配置：

节点数	3
操作系统	CentOS Linux release 7.5.1804 (Core)
内存	8GB

流程配置：

节点数	3
操作系统	CentOS Linux release 7.5.1804 (Core)
内存	16GB

注： 实际生产中按照需求分配内存，如果只是在vmvare中搭建虚拟机，内存可以调整为每台主机1-2GB即可

软件环境

软件	版本	下载地址
jdk	jdk-8u172-linux-x64	点击下载
zookeeeper	zookeeper-3.4.5-cdh5.14.2	点击下载
kafka	kafka_2.12-1.1.0	点击下载
kafka-manager	kafka-manager-1.3.3.17	源码下载

注： kafka-manager官方只提供源码下载，下载后编译：sbt clean dist，此编译需要x翻XX墙x才能完成，没有条件的话可以下载网友编译好的低一点的版本：kafka-manager-1.3.3.4

主机规划

3个节点角色规划如下：

主机名	CDHNode1	CDHNode2	CDHNode3
IP	192.168.223.201	192.168.223.202	192.168.223.203
zookeeper	yes	yes	yes
kafka	yes	yes	yes

注： ZooKeeper保持奇数个，如果需要高可用则不少于 3 个节点。具体原因，以后详叙。

主机安装前准备

1. 关闭所有节点的 SELinux

sed -i 's/^SELINUX=.*$/SELINUX=disabled/g' /etc/selinux/config 
setenforce 0

2. 关闭所有节点防火墙 firewalld or iptables

systemctl disable firewalld;  
systemctl stop firewalld;
systemctl disable iptables; systemctl stop iptables; 

3. 开启所有节点时间同步 ntpdate

echo "*/5 * * * * /usr/sbin/ntpdate asia.pool.ntp.org | logger -t NTP" >> /var/spool/cron/root

4. 设置所有节点语言编码以及时区

echo 'export TZ=Asia/Shanghai' >> /etc/profile
echo 'export LANG=en_US.UTF-8' >> /etc/profile . /etc/profile 

5. 所有节点添加kafka用户

useradd -m kafka
echo '123456' | passwd --stdin kafka
# 设置PS1
su - kafka
echo 'export PS1="\u@\h:\$PWD>"' >> ~/.bash_profile . ~/.bash_profile 

6. 设置kafka用户之间免密登录 首先在CDHNode1主机生成秘钥

su - kafka
ssh-keygen -t rsa	# 一直回车即可生成kafka用户的公钥和私钥
cd .ssh vi id_rsa.pub # 去掉私钥末尾的主机名 kafka@CDHNode1 cat id_rsa.pub > authorized_keys chmod 600 authorized_keys 

压缩.ssh文件夹

su - kafka
zip -r ssh.zip .ssh

随后分发ssh.zip到CDHNode2-3主机kafka用户家目录解压即完成免密登录

7. 主机内核参数优化以及最大文件打开数、最大进程数等参数优化 不同主机优化参数有可能不一样，故这里不作出具体优化方法，但如果kafka环境用于正式生产，必须优化，linux默认参数可能会导致kafka集群性能低下。

注： 以上操作需要使用 root 用户，到目前为止操作系统环境已经准备完成，以下开始正式安装，后面的操作如果不做特殊说明均使用 kafka 用户

安装jdk1.8

所有节点都需要安装，安装方式都一样 解压 jdk-8u172-linux-x64.tar.gz

tar zxvf jdk-8u172-linux-x64.tar.gz
mkdir -p /home/kafka/app
mv jdk-8u172-linux-x64 /home/kafka/app/jdk rm -f jdk-8u172-linux-x64.tar.gz 

配置环境变量 vi ~/.bash_profile 添加以下内容：

#java
export JAVA_HOME=/home/kafka/app/jdk
export CLASSPATH=.:$JAVA_HOME/lib:$CLASSPATH export PATH=$PATH:$JAVA_HOME/bin:$JAVA_HOME/jre/bin 

加载环境变量

. ~/.bash_profile

查看是否安装成功 java -version

java version "1.8.0_172"
Java(TM) SE Runtime Environment (build 1.8.0_172-b11)
Java HotSpot(TM) 64-Bit Server VM (build 25.172-b11, mixed mode)

如果出现以上结果证明安装成功。

安装zookeeper

首先在CDHNode1上安装

解压 zookeeper-3.4.5-cdh5.14.2.tar.gz

tar zxvf zookeeper-3.4.5-cdh5.14.2.tar.gz
mv zookeeper-3.4.5-cdh5.14.2 /home/kafka/app/zookeeper rm -f zookeeper-3.4.5-cdh5.14.2.tar.gz 

设置环境变量 vi ~/.bash_profile 添加以下内容：

#zk
export ZOOKEEPER_HOME=/home/kafka/app/zookeeper
export PATH=$PATH:$ZOOKEEPER_HOME/bin 

加载环境变量

. ~/.bash_profile

添加配置文件 vi /home/kafka/app/zookeeper/conf/zoo.cfg 添加以下内容：

# The number of milliseconds of each tick  
tickTime=2000
# The number of ticks that the initial  
# synchronization phase can take  
initLimit=10 # The number of ticks that can pass between # sending a request and getting an acknowledgement syncLimit=5 # the directory where the snapshot is stored. # do not use /tmp for storage, /tmp here is just # example sakes. #数据文件目录与日志目录 dataDir=/home/kafka/data/zookeeper/zkdata dataLogDir=/home/kafka/data/zookeeper/zkdatalog # the port at which the clients will connect clientPort=2181 #server.服务编号=主机名称：Zookeeper不同节点之间同步和通信的端口：选举端口（选举leader） server.1=CDHNode1:2888:3888 server.2=CDHNode2:2888:3888 server.3=CDHNode3:2888:3888 # 节点变更时只需在此添加或者删除相应的节点（所有节点配置都需要修改），然后在启动新增或者停止删除的节点即可 # administrator guide before turning on autopurge. # # http://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_maintenance # # The number of snapshots to retain in dataDir #autopurge.snapRetainCount=3 # Purge task interval in hours # Set to "0" to disable auto purge feature #autopurge.purgeInterval=1 

创建所需目录

mkdir -p /home/kafka/data/zookeeper/zkdata
mkdir -p /home/kafka/data/zookeeper/zkdatalog
mkdir -p /home/kafka/app/zookeeper/logs 

添加myid vim /home/kafka/data/zookeeper/zkdata/myid，添加：

1

注： 此数字来源于zoo.cfg中配置 server.1=CDHNode1:2888:3888行server后面的1，故CDHNode2填写2，CDHNode3填写3

配置日志目录 vim /home/kafka/app/zookeeper/libexec/zkEnv.sh ，修改以下参数为：

ZOO_LOG_DIR="$ZOOKEEPER_HOME/logs"
ZOO_LOG4J_PROP="INFO,ROLLINGFILE"

注： /home/kafka/app/zookeeper/libexec/zkEnv.sh 与 /home/kafka/app/zookeeper/bin/zkEnv.sh 文件内容相同。启动脚本 /home/kafka/app/zookeeper/bin/zkServer.sh 会优先读取/home/kafka/app/zookeeper/libexec/zkEnv.sh，当其不存在时才会读取 /home/kafka/app/zookeeper/bin/zkEnv.sh。

vim /home/kafka/app/zookeeper/conf/log4j.properties ，修改以下参数为：

zookeeper.root.logger=INFO, ROLLINGFILE
zookeeper.log.dir=/home/kafka/app/zookeeper/logs
log4j.appender.ROLLINGFILE=org.apache.log4j.RollingFileAppender 

复制zookeeper到CDHNode2-3

scp ~/.bash_profile CDHNode2:/home/kafka
scp ~/.bash_profile CDHNode3:/home/kafka
scp -pr /home/kafka/app/zookeeper CDHNode2:/home/kafka/app scp -pr /home/kafka/app/zookeeper CDHNode3:/home/kafka/app ssh CDHNode2 "mkdir -p /home/kafka/data/zookeeper/zkdata;mkdir -p /home/kafka/data/zookeeper/zkdatalog;mkdir -p /home/kafka/app/zookeeper/logs" ssh CDHNode2 "echo 2 > /home/kafka/data/zookeeper/zkdata/myid" ssh CDHNode3 "mkdir -p /home/kafka/data/zookeeper/zkdata;mkdir -p /home/kafka/data/zookeeper/zkdatalog;mkdir -p /home/kafka/app/zookeeper/logs" ssh CDHNode3 "echo 3 > /home/kafka/data/zookeeper/zkdata/myid" 

启动zookeeper 3个节点均启动

/home/kafka/app/zookeeper/bin/zkServer.sh start

查看节点状态

/home/kafka/app/zookeeper/bin/zkServer.sh status

如果一个节点为leader，另2个节点为follower，则说明Zookeeper安装成功

查看进程

jps

其中 QuorumPeerMain 进程为zookeeper

停止zookeeper

/home/kafka/app/zookeeper/bin/zkServer.sh stop

安装kafka

首先在CDHNode1上安装

解压 kafka_2.12-1.1.0.tgz

tar zxvf kafka_2.12-1.1.0.tgz
mv kafka_2.12-1.1.0 /home/kafka/app/kafka rm -f kafka_2.12-1.1.0.tgz 

设置环境变量 vi ~/.bash_profile 添加以下内容：

#kafka
export KAFKA_HOME=/home/kafka/app/kafka
export PATH=$PATH:$KAFKA_HOME/bin 

加载环境变量

. ~/.bash_profile

添加配置文件 vi /home/kafka/app/kafka/config/server.properties ：

# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements.  See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with # the License. You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # see kafka.server.KafkaConfig for additional details and defaults ############################# Server Basics ############################# # The id of the broker. This must be set to a unique integer for each broker. broker.id=0 #当前机器在集群中的唯一标识，和zookeeper的myid性质一样，CDHNode1为0,CDHNode1为2,CDHNode3为2 ############################# Socket Server Settings ############################# # The address the socket server listens on. It will get the value returned from # java.net.InetAddress.getCanonicalHostName() if not configured. # FORMAT: # listeners = listener_name://host_name:port # EXAMPLE: # listeners = PLAINTEXT://your.host.name:9092 #listeners=PLAINTEXT://:9092 port=9092 #当前kafka对外提供服务的端口默认是9092 # Hostname and port the broker will advertise to producers and consumers. If not set, # it uses the value for "listeners" if configured. Otherwise, it will use the value # returned from java.net.InetAddress.getCanonicalHostName(). #advertised.listeners=PLAINTEXT://your.host.name:9092 # Maps listener names to security protocols, the default is for them to be the same. See the config documentation for more details #listener.security.protocol.map=PLAINTEXT:PLAINTEXT,SSL:SSL,SASL_PLAINTEXT:SASL_PLAINTEXT,SASL_SSL:SASL_SSL # The number of threads that the server uses for receiving requests from the network and sending responses to the network num.network.threads=3 #这个是borker进行网络处理的线程数 # The number of threads that the server uses for processing requests, which may include disk I/O num.io.threads=8 #这个是borker进行I/O处理的线程数 # The send buffer (SO_SNDBUF) used by the socket server socket.send.buffer.bytes=102400 #发送缓冲区buffer大小，数据不是一下子就发送的，先回存储到缓冲区了到达一定的大小后在发送，能提高性能 # The receive buffer (SO_RCVBUF) used by the socket server socket.receive.buffer.bytes=102400 #kafka接收缓冲区大小，当数据到达一定大小后在序列化到磁盘 # The maximum size of a request that the socket server will accept (protection against OOM) socket.request.max.bytes=104857600 #这个参数是向kafka请求消息或者向kafka发送消息的请请求的最大数，这个值不能超过java的堆栈大小 ############################# Log Basics ############################# # A comma separated list of directories under which to store log files log.dirs=/home/kafka/data/kafka/kafka-logs #消息存放的目录，这个目录可以配置为“,”逗号分割的表达式，上面的num.io.threads要大于这个目录的个数这个目录 #如果配置多个目录，新创建的topic他把消息持久化的地方是，当前以逗号分割的目录中，那个分区数最少就放那一个 # The default number of log partitions per topic. More partitions allow greater # parallelism for consumption, but this will also result in more files across # the brokers. num.partitions=3 #默认的分区数，一个topic默认1个分区数 # The number of threads per data directory to be used for log recovery at startup and flushing at shutdown. # This value is recommended to be increased for installations with data dirs located in RAID array. num.recovery.threads.per.data.dir=1 ############################# Internal Topic Settings ############################# # The replication factor for the group metadata internal topics "__consumer_offsets" and "__transaction_state" # For anything other than development testing, a value greater than 1 is recommended for to ensure availability such as 3. offsets.topic.replication.factor=1 transaction.state.log.replication.factor=1 transaction.state.log.min.isr=1 ############################# Log Flush Policy ############################# # Messages are immediately written to the filesystem but by default we only fsync() to sync # the OS cache lazily. The following configurations control the flush of data to disk. # There are a few important trade-offs here: # 1. Durability: Unflushed data may be lost if you are not using replication. # 2. Latency: Very large flush intervals may lead to latency spikes when the flush does occur as there will be a lot of data to flush. # 3. Throughput: The flush is generally the most expensive operation, and a small flush interval may lead to excessive seeks. # The settings below allow one to configure the flush policy to flush data after a period of time or # every N messages (or both). This can be done globally and overridden on a per-topic basis. # The number of messages to accept before forcing a flush of data to disk #log.flush.interval.messages=10000 # The maximum amount of time a message can sit in a log before we force a flush #log.flush.interval.ms=1000 ############################# Log Retention Policy ############################# # The following configurations control the disposal of log segments. The policy can # be set to delete segments after a period of time, or after a given size has accumulated. # A segment will be deleted whenever *either* of these criteria are met. Deletion always happens # from the end of the log. # The minimum age of a log file to be eligible for deletion due to age log.retention.hours=168 #默认消息的最大持久化时间，168小时，7天 # A size-based retention policy for logs. Segments are pruned from the log unless the remaining # segments drop below log.retention.bytes. Functions independently of log.retention.hours. log.retention.bytes=1073741824 #日志数据存储的最大字节数,1GB # log.retention.hours 与 log.retention.bytes无论哪个先达到都会触发 # The maximum size of a log segment file. When this size is reached a new log segment will be created. log.segment.bytes=1073741824 #这个参数是：因为kafka的消息是以追加的形式落地到文件，当超过这个值的时候，kafka会新起一个文件 # The interval at which log segments are checked to see if they can be deleted according # to the retention policies log.retention.check.interval.ms=300000 message.max.byte=5242880 #消息保存的最大值5M default.replication.factor=3 #kafka保存消息的副本数，如果一个副本失效了，另一个还可以继续提供服务,必须小于等于集群节点数 replica.fetch.max.bytes=5242880 #取消息的最大直接数 ############################# Zookeeper ############################# # Zookeeper connection string (see zookeeper docs for details). # This is a comma separated host:port pairs, each corresponding to a zk # server. e.g. "127.0.0.1:3000,127.0.0.1:3001,127.0.0.1:3002". # You can also append an optional chroot string to the urls to specify the # root directory for all kafka znodes. zookeeper.connect=CDHNode1:2181,CDHNode2:2181,CDHNode3:2181 # Timeout in ms for connecting to zookeeper zookeeper.connection.timeout.ms=10000 ############################# Group Coordinator Settings ############################# # The following configuration specifies the time, in milliseconds, that the GroupCoordinator will delay the initial consumer rebalance. # The rebalance will be further delayed by the value of group.initial.rebalance.delay.ms as new members join the group, up to a maximum of max.poll.interval.ms. # The default value for this is 3 seconds. # We override this to 0 here as it makes for a better out-of-the-box experience for development and testing. # However, in production environments the default value of 3 seconds is more suitable as this will help to avoid unnecessary, and potentially expensive, rebalances during application startup. group.initial.rebalance.delay.ms=0 

创建所需目录

mkdir -p /home/kafka/data/kafka/kafka-logs

复制kafka到CDHNode2-3

scp ~/.bash_profile CDHNode2:/home/kafka
scp ~/.bash_profile CDHNode3:/home/kafka
scp -pr /home/kafka/app/kafka CDHNode2:/home/kafka/app scp -pr /home/kafka/app/kafka CDHNode3:/home/kafka/app ssh CDHNode2 "mkdir -p /home/kafka/data/kafka/kafka-logs" ssh CDHNode3 "mkdir -p /home/kafka/data/kafka/kafka-logs" 

修改 /home/kafka/app/kafka/config/server.properties 中的 broker.id CDHNode2为：1 ，CDHNode3为：2

启动kafka 3个节点均启动

/home/kafka/app/kafka/bin/kafka-server-start.sh -daemon /home/kafka/app/kafka/config/server.properties 

• -daemon 后台服务的方式启动

查看进程

jps

其中 Kafka 进程为zookeeper

停止kafka

/home/kafka/app/kafka/bin/kafka-server-stop.sh

kafka基本操作

创建topic

/home/kafka/app/kafka/bin/kafka-topics.sh --zookeeper CDHNode1:2181,CDHNode2:2181,CDHNode3:2181 --create --topic topic1 --replication-factor 3 --partitions 3 

• --replication-factor 指定副本数，必须小于等于集群节点数，否则报错
• --partitions 指定分区数，如果设置3个分区，则最多3个消费者同时消费信息

查看topic

/home/kafka/app/kafka/bin/kafka-topics.sh --zookeeper CDHNode1:2181,CDHNode2:2181,CDHNode3:2181 --describe --topic topic1 

• Leader：负责处理消息的读和写，Leader是从所有节点中随机选择的
• Replicas：列出了所有的副本节点，不管节点是否在服务中。
• Isr：是正在服务中的节点

删除topic

/home/kafka/app/kafka/bin/kafka-topics.sh --zookeeper CDHNode1:2181,CDHNode2:2181,CDHNode3:2181 --delete --topic topic1 

生成消息

使用kafka自带工具

/home/kafka/app/kafka/bin/kafka-console-producer.sh --broker-list CDHNode1:9092,CDHNode2:9092,CDHNode3:9092 --topic topic1 

进入控制台输入消息然后回车就生成一条消息

消费消息

使用kafka自带工具 第一种方法：

/home/kafka/app/kafka/bin/kafka-console-consumer.sh --zookeeper CDHNode1:2181,CDHNode2:2181,CDHNode3:2181 --from-beginning --topic topic1 

不会记录消费者offset，每次都是从头开始

第二种方法：

/home/kafka/app/kafka/bin/kafka-console-consumer.sh --zookeeper CDHNode1:2181,CDHNode2:2181,CDHNode3:2181 --group test_group --topic topic1 

指定消费者组，会记录消费者offset

注： kafka中数据的删除跟有没有消费者消费完全无关。数据的删除，只跟kafka broker的这两个配置有关：

log.retention.hours=48 #数据最多保存48小时 log.retention.bytes=1073741824 #数据最多1G 

查看消费者组消费信息

/home/kafka/app/kafka/bin/kafka-consumer-groups.sh --zookeeper CDHNode1:2181 --group test_group --describe 

安装管理监控工具Kafka-Manager

CDHNode1上解压安装下载好的 kafka-manager-1.3.3.4.zip

unzip kafka-manager-1.3.3.4.zip
mv kafka-manager-1.3.3.4 /home/kafka/app/kafka-manager mkdir -p /home/kafka/app/kafka-manager/logs rm -f kafka-manager-1.3.3.4.zip 

添加配置文件 vi /home/kafka/app/kafka-manager/conf/application.cof：

# Copyright 2015 Yahoo Inc. Licensed under the Apache License, Version 2.0
# See accompanying LICENSE file.

# This is the main configuration file for the application.
# ~~~~~

# Secret key # ~~~~~ # The secret key is used to secure cryptographics functions. # If you deploy your application to several instances be sure to use the same key! play.crypto.secret="^<csmm5Fx4d=r2HEX8pelM3iBkFVv?k[mc;IZE<_Qoq8EkX_/7@Zt6dP05Pzea3U" play.crypto.secret=${?APPLICATION_SECRET} # The application languages # ~~~~~ play.i18n.langs=["en"] play.http.requestHandler = "play.http.DefaultHttpRequestHandler" play.http.context = "/" play.application.loader=loader.KafkaManagerLoader kafka-manager.zkhosts="CDHNode1:2181,CDHNode2:2181,CDHNode3:2181" kafka-manager.zkhosts=${?ZK_HOSTS} pinned-dispatcher.type="PinnedDispatcher" pinned-dispatcher.executor="thread-pool-executor" application.features=["KMClusterManagerFeature","KMTopicManagerFeature","KMPreferredReplicaElectionFeature","KMReassignPartitionsFeature"] akka { loggers = ["akka.event.slf4j.Slf4jLogger"] loglevel = "INFO" } # 开启http基本密码验证 basicAuthentication.enabled=true basicAuthentication.username="admin" basicAuthentication.password="password" basicAuthentication.realm="Kafka-Manager" kafka-manager.consumer.properties.file=${?CONSUMER_PROPERTIES_FILE} 

启动（nohup）

nohup /home/kafka/app/kafka-manager/bin/kafka-manager -Dconfig.file=/home/kafka/app/kafka-manager/conf/application.conf -Dhttp.port=8080 > /home/kafka/app/kafka-manager/logs/server.log 2>&1 & 

• -Dhttp.port=8080 指定端口，默认9000
• -Dconfig.file 指定配置文件

停止

kill pid

访问

http://CDHNode1:8080

点击【Cluster】>【Add Cluster】打开如下添加集群的配置界面： 自定义集群名称和zookeeper地址 版本没有1.1.0，选择最新的0.10.1.0即可

其他broker的配置可以根据自己需要进行配置，默认情况下，点击【保存】时，会提示几个默认值为1的配置错误，需要配置为>=2的值。保存成功后，点击【Go to cluster view.】打开当前的集群界面。

关于kafka-manager的其他功能和操作可以参考官网：https://github.com/yahoo/kafka-manager。

开启kafka-server JMX 开启后kafka-manager能够显示更多信息，后期补充。