redis压力测试详解

时间:2022-04-12 11:31:35

redis做压测可以用自带的redis-benchmark工具,使用简单,效果也比较不错。

 

linux下一般无需下载,windows下redis-benchmark压力测试工具下载地址:http://www.daixiaorui.com/source/18.html(解压后的redis-benchmark.exe)

 

压测命令:redis-benchmark -h 127.0.0.1 -p 6379 -c 50 -n 10000

 

-c表示连接数

-n表示请求数

更多参数请输入 --help 查看~

 

压测需要一段时间,因为它需要依次压测多个命令的结果,如:get、set、incr、lpush等等,所以我们需要耐心等待,如果只需要压测某个命令,如:get,那么可以在以上的命令后加一个参数-t(红色部分):

redis-benchmark -h 127.0.0.1 -p 6086 -c 50 -n 10000 -t get

 

压测结果:

[root@1234 ~]# redis-benchmark -h 127.0.0.1 -p 6086 -c 50 -n 10000 -t get

====== GET ======

  10000 requests completed in 0.15 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.16% <= 1 milliseconds

100.00% <= 1 milliseconds

68027.21 requests per second

 

不带-t的压测结果如下:

[root@1234 ~]# redis-benchmark -h 127.0.0.1 -p 6086 -c 50 -n 10000

====== PING_INLINE ======

  10000 requests completed in 0.15 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.35% <= 1 milliseconds

100.00% <= 1 milliseconds

67114.09 requests per second

 

====== PING_BULK ======

  10000 requests completed in 0.15 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.38% <= 1 milliseconds

100.00% <= 1 milliseconds

66666.66 requests per second

 

====== SET ======  (处理set的性能)

  10000 requests completed in 0.17 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.11% <= 1 milliseconds

99.51% <= 25 milliseconds

100.00% <= 25 milliseconds

57142.86 requests per second

 

====== GET ======  (处理get请求的性能)

  10000 requests completed in 0.15 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.09% <= 1 milliseconds

99.51% <= 11 milliseconds

100.00% <= 12 milliseconds

66666.66 requests per second

 

====== INCR ======

  10000 requests completed in 0.15 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.33% <= 1 milliseconds

100.00% <= 1 milliseconds

66666.66 requests per second

 

====== LPUSH ======

  10000 requests completed in 0.16 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.20% <= 1 milliseconds

100.00% <= 1 milliseconds

64516.13 requests per second

 

====== LPOP ======

  10000 requests completed in 0.16 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

98.56% <= 1 milliseconds

99.51% <= 14 milliseconds

100.00% <= 14 milliseconds

61349.69 requests per second

 

====== SADD ======

  10000 requests completed in 0.15 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.36% <= 1 milliseconds

100.00% <= 1 milliseconds

67114.09 requests per second

 

====== SPOP ======

  10000 requests completed in 0.14 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.32% <= 1 milliseconds

100.00% <= 1 milliseconds

69930.07 requests per second

 

====== LPUSH (needed to benchmark LRANGE) ======

  10000 requests completed in 0.15 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.47% <= 1 milliseconds

100.00% <= 1 milliseconds

67567.57 requests per second

 

====== LRANGE_100 (first 100 elements) ======

  10000 requests completed in 0.14 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.41% <= 1 milliseconds

100.00% <= 1 milliseconds

72992.70 requests per second

 

====== LRANGE_300 (first 300 elements) ======

  10000 requests completed in 0.14 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.41% <= 1 milliseconds

100.00% <= 1 milliseconds

72463.77 requests per second

 

====== LRANGE_500 (first 450 elements) ======

  10000 requests completed in 0.14 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.35% <= 1 milliseconds

100.00% <= 1 milliseconds

68965.52 requests per second

 

====== LRANGE_600 (first 600 elements) ======

  10000 requests completed in 0.15 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

99.37% <= 1 milliseconds

100.00% <= 1 milliseconds

66225.17 requests per second

 

====== MSET (10 keys) ======

  10000 requests completed in 0.17 seconds

  50 parallel clients

  3 bytes payload

  keep alive: 1

 

94.94% <= 1 milliseconds

100.00% <= 1 milliseconds

58479.53 requests per second

 

这样看起来很冗余,如果我们只想看最终的结果,可以带上参数-q,完整的命令如下:

redis-benchmark -h 127.0.0.1 -p 6086 -c 50 -n 10000 -q

 

显示结果为:

 

[root@1234 ~]# redis-benchmark -h 127.0.0.1 -p 6086 -c 50 -n 10000 -q

PING_INLINE: 66225.17 requests per second

PING_BULK: 65789.48 requests per second

SET: 66666.66 requests per second

GET: 69444.45 requests per second

INCR: 62893.08 requests per second

LPUSH: 65789.48 requests per second

LPOP: 68027.21 requests per second

SADD: 64935.07 requests per second

SPOP: 67114.09 requests per second

LPUSH (needed to benchmark LRANGE): 62893.08 requests per second

LRANGE_100 (first 100 elements): 69444.45 requests per second

LRANGE_300 (first 300 elements): 68965.52 requests per second

LRANGE_500 (first 450 elements): 68965.52 requests per second

LRANGE_600 (first 600 elements): 68965.52 requests per second

MSET (10 keys): 59171.60 requests per second

 

温馨提示:

压测结果跟机器的性能有关,其中windows下压测的结果要比在linux下差一大截。