redis(1.21)redis6.0安装
作者:互联网
【1】一键安装(基于集群)
如果要装单实例,请注释掉集群参数
#!/bin/bash source /etc/profile usage(){ echo $1 echo '$#': $# echo "for example: sh install_redis.sh -f redis-6.2.4.tar.gz -p 6380 -d /data/redis/ -h usage -f redis file tar.gz -d datadir -p port " } while getopts :f:d:p:h opt do case "$opt" in h) usage && exit ;; f) redis_file=$OPTARG ;; p) port=$OPTARG ;; d) datadir=$OPTARG ;; *) usage && exit;; esac done if [ ! $port ];then port=6379 fi if [ `netstat -ano|grep ${port}|wc -l` -gt 0 ];then echo "the port have already been used:${port}" netstat -ano|grep ${port} exit fi echo -n "check gcc......" gcc -v >/dev/null 2>&1 if [ ! $? -eq 0 ];then echo "None" echo "yum -y install gcc 1>/dev/null 2>gcc_error.log......" yum -y install gcc 1>/dev/null 2>gcc_error.log #yum -y install tcl fi echo "ok" if [ ! $datadir ];then datadir=/data/redis/ fi if [ ! $redis_file ] || [ ! -f $redis_file ];then echo 'redis file option not input' usage && exit fi #------------------ redis安装 --------------------- mkdir -p ${datadir}${port}_redis dir=`echo ${redis_file} |awk -F'.tar' '{print $1}'` if [ ! -d $dir ];then tar -zxf $redis_file dir=`echo ${redis_file} |awk -F'.tar' '{print $1}'` cd $dir echo -n " making redis....... make 1>/dev/null 2>make_error.log" make 1>/dev/null 2>make_error.log if [ ! $? -eq 0 ];then echo "make error" exit fi # make MALLOC=libc echo "ok" cd ./src make PREFIX=$datadir install elif [ ! -f "${dir}/src/redis-cli" ];then cd $dir make if [ ! $? -eq 0 ];then echo "make error" exit fi make PREFIX=$datadir install # make MALLOC=libc elif [ ! -f "${datadir}bin/redis-cli" ];then cd $dir make PREFIX=$datadir install fi if [ `echo ${PATH}|grep ${datadir}bin|wc -l` -eq 0 ];then echo "export PATH=${PATH}:/data/redis/bin" >> /etc/profile source /etc/profile fi if [ -f "${datadir}${port}_redis/${port}_redis.conf" ];then echo "redis config file:${datadir}${port}_redis/${port}_redis.conf is already exists!Please check the file info." exit fi #------------配置文件 /data/redis/redis_6379/redis_6379.conf---------- cat <<eof>>${datadir}${port}_redis/${port}_redis.conf #network protected-mode no port ${port} #general daemonize yes pidfile ${datadir}${port}_redis/${port}_redis.pid supervised no loglevel notice logfile "${datadir}${port}_redis/${port}_redis.log" syslog-enabled no databases 16 #SNAPSHOTTING save 86400 1 stop-writes-on-bgsave-error no rdbcompression yes dbfilename dump.rdb dir ${datadir}${port}_redis/ requirepass 123456 masterauth 123456 #rename-command FLUSHALL 1123123_FLUSHALL #rename-command FLUSHDB 1124124_FLUSHDB #rename-command CONFIG 1123123_CONFIG #replication #slaveof 180.153.250.135 ${port} #slave-read-only yes repl-diskless-sync no repl-diskless-sync-delay 5 repl-disable-tcp-nodelay no slave-priority 100 #io thread io-threads 4 # 开启 4 个 IO 线程 io-threads-do-reads yes #读请求解析也是用 IO 线程 #limit maxclients 10000 maxmemory 100gb maxmemory-policy allkeys-lru #append only mode appendonly yes appendfsync no no-appendfsync-on-rewrite no auto-aof-rewrite-percentage 100 #cluster cluster-enabled yes #-- 允许实例加入集群 cluster-config-file nodes-${port}.conf #-- 每个加入群集的实例都会有一个 cluster-node-timeout 15000 #-- ms,节点响应超时 cluster-replica-validity-factor 10 #-- 这个系数是 超时事件* 它 + 集群心跳检查时间; cluster-require-full-coverage yes # --当为no时: 如果请求过来发现集群无法处理(比如某个主实例挂了又没有副本顶上,造成缺失部分hash槽),则可以重新分配hash槽 覆盖原有数据; yes 则必须所有hash槽 ok 状态集群才能访问 cluster-migration-barrier 1 # -- 迁移壁垒,即当某个主实例下没有副本,同时又有某个主实例下有超过1个以上的副本,则至少保留一个副本,多的则可以迁移到没有副本的实例下做副本 cluster-replica-no-failover no # -- 副本是否故障转移,如果为 yes 则副本无法顶替主实例自动故障转移(可以手动) cluster-allow-reads-when-down no # --6.0参数,当集群不可用时,是否允许节点读写 eof #--------------配置、启动、连接、核验------------- redis-server ${datadir}${port}_redis/${port}_redis.conf sleep 3 ps -ef|grep redis-server|grep -v grep if [ `cat /etc/rc.local|grep "redis-server ${datadir}${port}_redis/${port}_redis.conf"|wc -l` -eq 0 ];then echo "redis-server ${datadir}${port}_redis/${port}_redis.conf">>/etc/rc.local fi
标签:redis6.0,no,make,redis,echo,datadir,1.21,port 来源: https://www.cnblogs.com/gered/p/15018760.html