测试环境

1, 操作系统

centos7

内核:3.10.0-123.20.1.el7.x86_64

2, 内存:

grep MemTotal /proc/meminfo

MemTotal:        7138276 kB

3, 请求数 10000,并发clients 50

4,服务器版本信息

   1,redis:   Redis server v=3.2.1 sha=00000000:0 malloc=libc bits=64 build=939341c443eba189

   2,ssdb:  1.9.4

redis-benchmark -n 10000 -c 50

====== PING_INLINE ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1

31746.03 requests per second

====== PING_BULK ======
10000 requests completed in 0.29 seconds
3 bytes payload
keep alive: 1

34246.57 requests per second

====== SET ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1

32679.74 requests per second

====== GET ======
10000 requests completed in 0.30 seconds
3 bytes payload
keep alive: 1
33222.59 requests per second

====== INCR ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1
32051.28 requests per second

====== LPUSH ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1
32679.74 requests per second

====== RPUSH ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1

32154.34 requests per second

====== LPOP ======
10000 requests completed in 0.32 seconds
3 bytes payload
keep alive: 1

31152.65 requests per second

====== RPOP ======
10000 requests completed in 0.29 seconds
3 bytes payload
keep alive: 1
34843.21 requests per second

====== SADD ======
10000 requests completed in 0.32 seconds
3 bytes payload
keep alive: 1

31347.96 requests per second

====== SPOP ======
10000 requests completed in 0.30 seconds
3 bytes payload
keep alive: 1

33557.05 requests per second

====== LPUSH (needed to benchmark LRANGE) ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1

31847.13 requests per second

====== LRANGE_100 (first 100 elements) ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1

32258.06 requests per second

====== LRANGE_300 (first 300 elements) ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1

32679.74 requests per second

====== LRANGE_500 (first 450 elements) ======
10000 requests completed in 0.33 seconds
3 bytes payload
keep alive: 1

30303.03 requests per second

====== LRANGE_600 (first 600 elements) ======
10000 requests completed in 0.31 seconds
3 bytes payload
keep alive: 1

32362.46 requests per second

====== MSET (10 keys) ======
10000 requests completed in 0.32 seconds
3 bytes payload keep alive: 1

30959.75 requests per second

ssdb压测

========== set ==========
qps: 14691, time: 0.681 s
========== get ==========
qps: 20075, time: 0.498 s
========== del ==========
qps: 12913, time: 0.774 s
========== hset ==========
qps: 13218, time: 0.757 s
========== hget ==========
qps: 18534, time: 0.540 s
========== hdel ==========
qps: 12697, time: 0.788 s
========== zset ==========
qps: 8334, time: 1.200 s
========== zget ==========
qps: 17628, time: 0.567 s ========== zdel ==========
qps: 11980, time: 0.835 s
========== qpush ==========
qps: 8499, time: 1.177 s
========== qpop ==========
qps: 10060, time: 0.994 s

CPU 对比:

   redis:     百分之十几到 七十几         

   ssdb:    15%-25%

内存对比:

  redis:   0.0%

  ssdb:  1%左右浮动  +-0.5
   1,redis比较消耗cpu资源
   2,ssdb cpu 内存资源较为稳定
   3,redis    ssdb查询优于redis , 更新略逊redis
  
     ssdb基于磁盘存储,对于想要redis的查询性能,对写入性能不敏感,且担心内存溢出,可以考虑尝试ssdb


blog comments powered by Disqus

Published

2016-08-25

Categories


Tags