Zookeeper安装与调优部署参考文档(Linux)

时间 2021-01-29

标签 html java linux c++ apache 安全 bash 服务器 cookie tcp 栏目 Zookeeper 繁體版

原文原文链接

1. 安装环境准备

1.1 主机环境准备

1.1.1. 关闭selinux

sed -i 's/SELINUX=enforcing/SELINUX=disabled/g' /etc/selinux/config
setenforce 0

1.1.2. 软件下载

apache-zookeeper-3.6.1-bin.tar.gz：下载地址html

1.1.3. 部署规划

软件安装路径 /usr/local/zookeeper
端口规划 2192java

1.1.4. 系统主机时间、时区、系统语言

 本节视实际状况须要操做
 修改时区linux

ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime

 修改系统语言环境c++

echo 'LANG="en_US.UTF-8"' >> /etc/profile && source /etc/profile

 配置主机NTP时间同步apache

yum -y install ntp
systemctl enable ntpd && systemctl start ntpd
echo 'server ntp1.aliyun.com' >> /etc/ntp.conf
echo 'server ntp2.aliyun.com' >> /etc/ntp.conf

2. Zookeeper安装部署

2.1 Zookeeper依赖安装及部署

 添加用户与用户组（用户名请自行定义）安全

groupadd -r middleware && useradd -s /sbin/nologin -r -M -g middleware middleware

 JDK安装部署bash

tar -zxvf jdk-8u231-linux-x64.tar.gz -C /usr/local/
cat >>/etc/profile<<EOF
export JAVA_HOME=/usr/local/jdk1.8.0_231
export JRE_HOME=\${JAVA_HOME}/jre
export CLASSPATH=.:\${JAVA_HOME}/lib:\${JRE_HOME}/lib
export PATH=\${JAVA_HOME}/bin:\$PATH
EOF
source /etc/profile
java -version

 下载apache-zookeeper-3.6.1-bin.tar.gz安装包，并解压安装服务器

yum -y install gcc gcc-c++ automake autoconf libevent-devel libevent make wget net-tools
cd /opt
wget https://mirror.bit.edu.cn/apache/zookeeper/zookeeper-3.6.1/apache-zookeeper-3.6.1-bin.tar.gz
tar -zxvf apache-zookeeper-3.6.1-bin.tar.gz -C /usr/local/
cd /usr/local/
mv apache-zookeeper-3.6.1-bin zookeeper
mkdir -p zookeeper/data/zookeeper
mkdir zookeeper/dataLog
cd zookeeper/conf
cp zoo_sample.cfg zoo.cfg

 修改zookeeper数据存储路径与链接端口cookie

vi zoo.cfg
dataDir=/usr/local/zookeeper/data/zookeeper
dataLogDir=/usr/local/zookeeper/dataLog
clientPort=2192
chown -R middleware:middleware /usr/local/zookeeper

 配置Zookeeper环境变量tcp

cat >>/etc/profile<< EOF
export PATH="\$PATH:/usr/local/zookeeper/bin"
EOF
source /etc/profile

2.2 配置zookeeper系统服务

2.2.1. 针对6系统添加系统服务

一、添加防火墙策略
（1）全部机器可访问

iptables -A INPUT -p tcp --dport 2192 -j ACCEPT
service iptables save

（2）特定IP192.168.31.130可访问本机2192端口

iptables -A INPUT -p tcp -s 192.168.31.130 --dport 2192 -j ACCEPT
service iptables save

二、添加zookeeper系统服务启动脚本

cd /usr/local/zookeeper/bin/
sed -i '77aJAVA_HOME="/usr/local/jdk1.8.0_231"' zkEnv.sh
vi /etc/init.d/zookeeper
#!/bin/bash
#
# zookeeper  start/stop the zookeeper daemon
#
# chkconfig: 345 80 20
# description: zookeeper is a message server.
#
ZOOKEEPER_HOME=/usr/local/zookeeper
PIDFILE=/usr/local/zookeeper/data/zookeeper/zookeeper_server.pid

case $1 in
    start)
        if [ -f $PIDFILE ]
        then
                echo "$PIDFILE exists, process is already running"
        else
                echo "Starting zookeeper server..."
                sudo -u middleware $ZOOKEEPER_HOME/bin/zkServer.sh start
        fi
        ;;
    stop)
        if [ ! -f $PIDFILE ]
        then
                echo "$PIDFILE does not exist, process is not running"
        else
                sudo -u middleware $ZOOKEEPER_HOME/bin/zkServer.sh stop
        fi
        ;;
    status)
        if [ ! -f $PIDFILE ]
        then
                echo "$PIDFILE does not exist, process is not running"
        else
                sudo -u middleware $ZOOKEEPER_HOME/bin/zkServer.sh status
                echo "Zookeeper service is running..."
        fi
        ;;
    restart)
        sudo -u middleware $ZOOKEEPER_HOME/bin/zkServer.sh restart
        ;;
    *)
        echo "Please use start|stop|status|restart as first argument"
        ;;
esac

三、配置zookeeper系统服务及自启动

chmod +x /etc/init.d/zookeeper
chkconfig --add zookeeper && chkconfig zookeeper on
chkconfig --list zookeeper

四、启动与中止zookeeper服务

service zookeeper start
ps -ef|grep zookeeper
service zookeeper stop

2.2.2. 针对7系统添加系统服务

一、添加防火墙策略
（1）全部机器可访问

firewall-cmd --permanent --zone=public --add-port=2192/tcp
firewall-cmd --reload

（2）特定IP192.168.31.130可访问本机2192端口

firewall-cmd --permanent --add-rich-rule="rule family="ipv4" source address="192.168.31.130" port protocol="tcp" port="2192" accept"
firewall-cmd --reload

（3）特定IP段192.168.142.0/24可访问本机2192端口

firewall-cmd --permanent --add-rich-rule="rule family="ipv4" source address="192.168.142.0/24" port protocol="tcp" port="2192" accept"
firewall-cmd --reload

二、添加zookeeper系统服务启动脚本
获取当前服务器PATH路径信息，并将此信息添加到zookeeper系统服务中

echo $PATH
/usr/local/jdk1.8.0_231/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin
cat >/usr/lib/systemd/system/zookeeper.service<<EOF
[Unit]
Description=Zookeeper
After=network.target

[Service]
Type=forking
Environment=ZOO_LOG_DIR=/usr/local/zookeeper/logs
Environment=PATH=/usr/local/jdk1.8.0_231/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/root/bin
PIDFile=/usr/local/zookeeper/data/zookeeper/zookeeper_server.pid
ExecStart=/usr/local/zookeeper/bin/zkServer.sh start
ExecStop=/usr/local/zookeeper/bin/zkServer.sh stop
ExecRestart=/usr/local/zookeeper/bin/zkServer.sh restart
User=middleware
Group=middleware

[Install]
WantedBy=multi-user.target
EOF

三、配置zookeeper系统服务及自启动

systemctl daemon-reload
systemctl enable zookeeper.service

四、启动与中止zookeeper服务

systemctl start zookeeper
ps -ef|grep zookeeper
systemctl stop zookeeper

3. Zookeeper加固

3.1 最小化权限用户启动

 用户名请自行定义

groupadd -r middleware && useradd -s /sbin/nologin -r -M -g middleware middleware

3.2 预防DOS***

 限制zookeeper客户端的最大链接数。

vi /usr/local/zookeeper/conf/zoo.cfg
maxClientCnxns=60

3.3 修改默认2181端口

 默认状况下，zookeeper默认使用2181端口，请修改默认监听端口，如本文档使用的是2192

vi /usr/local/zookeeper/conf/zoo.cfg
clientPort=2192

3.4 禁用管理控制台

 如不须要使用zookeeper的管理控制台，建议禁用（zookeeper的管理控制台是由jetty启动的，默认为http，存在必定的信息泄露及安全隐患。）
 操做指导：
在bin/zkServer.sh文件中，将以下

vi /usr/local/zookeeper/bin/zkServer.sh
start)
    echo  -n "Starting zookeeper ... "
    if [ -f "$ZOOPIDFILE" ]; then
      if kill -0 `cat "$ZOOPIDFILE"` > /dev/null 2>&1; then
         echo $command already running as process `cat "$ZOOPIDFILE"`.
         exit 1
      fi
    fi
    nohup "$JAVA" $ZOO_DATADIR_AUTOCREATE "-Dzookeeper.log.dir=${ZOO_LOG_DIR}" \
"-Dzookeeper.log.file=${ZOO_LOG_FILE}" "-Dzookeeper.root.logger=${ZOO_LOG4J_PROP}" \
修改成（即在nohup这一行，添加 "-Dzookeeper.admin.enableServer=false"）
start)
    echo  -n "Starting zookeeper ... "
    if [ -f "$ZOOPIDFILE" ]; then
      if kill -0 `cat "$ZOOPIDFILE"` > /dev/null 2>&1; then
         echo $command already running as process `cat "$ZOOPIDFILE"`.
         exit 1
      fi
    fi
    nohup "$JAVA" $ZOO_DATADIR_AUTOCREATE "-Dzookeeper.log.dir=${ZOO_LOG_DIR}" \
    "-Dzookeeper.log.file=${ZOO_LOG_FILE}" "-Dzookeeper.admin.enableServer=false" "-Dzookeeper.root.logger=${ZOO_LOG4J_PROP}" \

3.5 日志清理

 建议设置对zookeeper日志的按期清理功能，在配置文件中清理日志策略，以下所示：

vi /usr/local/zookeeper/conf/zoo.cfg
autopurge.snapRetainCount=10
autopurge.purgeInterval=24
参数说明：
autopurge.snapRetainCount=10  //保留多少个快照
autopurge.purgeInterval=24     //多少小时清理一次

3.6 配置事务日志与快照日志分离

vi /usr/local/zookeeper/conf/zoo.cfg
dataDir=/usr/local/zookeeper/data/zookeeper
dataLogDir=/usr/local/zookeeper/dataLog

3.7 添加对zookeeper的指定IP受权访问

 zookeeper在默认状况下，是容许任意客户端未经受权访问，存在很大的安全隐患。具体链接指令以下：

/usr/local/zookeeper/bin/zkCli.sh -server 127.0.0.1:2192
WatchedEvent state:SyncConnected type:None path:null    //敲回车

 等待输入操做指令，如建立用户、受权等

[zk: 127.0.0.1:2192(CONNECTED) 0]

 getAcl / 表示查看当前权限 quit 表示退出客户端链接

[zk: 127.0.0.1:2192(CONNECTED) 3] getAcl /
'world,'anyone
: cdrwa

 添加可访问IP，一组可访问ip间以符号,隔开，格式以下

[zk: 127.0.0.1:2192(CONNECTED) 3] 
setAcl / ip:192.168.31.130:cdrwa,ip:127.0.0.1:cdrwa

 查看权限是否添加成功

[zk: 127.0.0.1:2192(CONNECTED) 3] getAcl /
'ip,'192.168.31.130
: cdrwa
'ip,'127.0.0.1
: cdrwa

 回退方法

[zk: 127.0.0.1:2192(CONNECTED) 3] setAcl / world:anyone:cdrwa

 zookeeper身份的认证有4种方式：
(1)world：默认方式，至关于全世界都能访问
(2)auth：表明已经认证经过的用户(cli中能够经过addauth digest user:pwd 来添加当前上下文中的受权用户)
(3)digest：即用户名:密码这种方式认证，这也是业务系统中最经常使用的，用username:password 字符串来产生一个MD5串，而后该串被用来做为ACL ID，认证是经过明文发送username:password 来进行的，当用在ACL时，表达式为username:base64 ，base64是password的SHA1摘要的编码；
(4)ip：使用Ip地址认证

 ID受权对象ID是指，权限赋予的用户或者一个实体，例如：IP 地址或者机器，受权模式受权对象有：
(1)IP：一般是一个IP地址或IP段，例如“192.168.29.100”或“192.168.29.100/110”
(2)Digest：自定义，一般是“username:BASE64(SHA-1(username:password))”,例如"foo:kWN6aNsbjcKWpqjiV7cg0N24raU="
(3)Word 只有一个ID：“anyone”
(4)Super：与Digest模式一致
 zookeeper支持的权限有5种分别是（其中delete是指对子节点的删除权限，其它4种权限指对自身节点的操做权限）

cdrwa:
        create: 能够建立子节点；
        read: 能够获取节点数据以及当前节点的子节点列表；
        write: 能够为节点设置数据；
        delete: 能够删除子节点；
        admin: 能够为节点设置权限。

3.8 帐号与认证

一、经过zkCli.cmd 进入zookeeper客户端
/usr/local/zookeeper/bin/zkCli.sh -server 127.0.0.1:2192
WatchedEvent state:SyncConnected type:None path:null    //敲回车
二、使用auth方式加密，添加用户名crm和密码pwd
addauth digest crm:crm#pwd
三、授予/dubbo auth权限
setAcl /dubbo auth:crm:crm#pwd:rwadc
四、查看目录加密后的权限
getAcl /dubbo

3.9 配置防火墙策略

 根据操做系统的不一样，参考2.2章节（注意若是是配置特定IP地址访问时，也要添加3.7章节中添加的指定IP）

3.10 按期升级

 使用官方最新稳定版本

4. Zookeeper优化

4.1 优化内核参数

cat >>/etc/sysctl.conf<<EOF
fs.file-max = 6815744
net.ipv4.tcp_tw_reuse = 1 
net.ipv4.tcp_tw_recycle = 1 
net.ipv4.tcp_fin_timeout = 30 
net.ipv4.ip_local_port_range = 10000 65000 
net.ipv4.tcp_max_syn_backlog = 8192 
net.ipv4.tcp_max_tw_buckets = 10000
net.core.somaxconn=4000
net.ipv4.tcp_syncookies = 1
net.core.netdev_max_backlog = 262144
net.ipv4.tcp_max_orphans = 262144
EOF
sysctl -p

4.2 系统资源限制

cat >>/etc/security/limits.conf<<EOF
* soft nofile 65525
* hard nofile 65525
* soft nproc 65525
* hard nproc 65525
EOF