分享一个脚本
[root@ELK-proxy-0-253 ~]# cat benchmark.sh
#! /bin/bash
port=7617
ip=192.168.0.253
cli="/root/redis-5.0.5/src/redis-cli"
for ((i=0;i<100000;i++))
do
echo -en "helloworld" | ${cli} -h ${ip} -p ${port} -x set name${i} >> ./benchmark.log
done
跑脚本 && 检查数据
[root@ELK-proxy-0-253 ~]# bash benchmark.sh
[root@ELK-proxy-0-253 ~]# cd /data/redis/
============ bgsave 落盘 =================
[root@ELK-proxy-0-253 redis]# redis-cli -p 6379
127.0.0.1:6379> BGSAVE
Background saving started
127.0.0.1:6379>
[root@ELK-proxy-0-253 redis]# redis-cli -p 6380
127.0.0.1:6380> BGSAVE
Background saving started
127.0.0.1:6380>
[root@ELK-proxy-0-253 redis]# redis-cli -p 6381
127.0.0.1:6381> BGSAVE
Background saving started
127.0.0.1:6381>
[root@ELK-proxy-0-253 redis]# redis-cli -p 6382
127.0.0.1:6382> BGSAVE
Background saving started
127.0.0.1:6382>
[root@ELK-proxy-0-253 redis]# redis-cli -p 6383
127.0.0.1:6383> BGSAVE
Background saving started
127.0.0.1:6383>
================ 所有实例数据量大小一致=================
[root@ELK-proxy-0-253 redis]# du -s ./*
17736 ./6379
17736 ./6380
17736 ./6381
17736 ./6382
17736 ./6383
证明通过predixy代理之后，redis cluster在压力测试后，数据写入正常
性能分析
大部分set操作耗时在200毫秒左右
第1列代表小于等于第二列的耗时，第3列表示耗时总和，第四列表示总操作次数
T开头那一行，第二列表示平均耗时，第三列表示总耗时，第四列表示总操作次数
重启之后，统计数据全部归零。
# LatencyMonitor
LatencyMonitorName:all
<= 100 5837 72 0.07%
<= 200 12096998 93673 91.50%
<= 300 1829037 7969 99.28%
<= 400 189120 569 99.84%
<= 500 39051 89 99.92%
<= 600 15878 29 99.95%
<= 700 7617 12 99.96%
<= 800 5192 7 99.97%
<= 900 10312 12 99.98%
<= 1000 7758 8 99.99%
<= 1200 5418 5 99.99%
<= 1400 1273 1 100.00%
<= 2000 3883 2 100.00%
<= 2500 2376 1 100.00%
<= 3000 2950 1 100.00%
<= 3500 3372 1 100.00%
T 138 14226072 102451
LatencyMonitorName:get
LatencyMonitorName:set
<= 100 3532 36 0.04%
<= 200 12096896 93672 91.50%
<= 300 1829037 7969 99.28%
<= 400 189120 569 99.84%
<= 500 39051 89 99.92%
<= 600 15878 29 99.95%
<= 700 7617 12 99.96%
<= 800 5192 7 99.97%
<= 900 10312 12 99.98%
<= 1000 7758 8 99.99%
> 1000 19272 11 100.00%
T 138 14223665 102414
主机性能曲线图
下一篇 redis 集群性能提升篇