redis压力测试

redis自带的redis-benchmark工具

Redis 自带了一个叫redis-benchmark的工具来模拟 N 个客户端同时发出 M 个请求。 （类似于 Apache ab 程序）。你可以使用 redis-benchmark -h 来查看基准参数。

• 1 使用方法 redis-benchmark [-h ] [-p ] [-c ] [-n [-k ]

[root@redis-test-slave ~ ]$ redis-benchmark --help
Usage: redis-benchmark [-h ] [-p ] [-c ] [-n [-k ]
-h Server hostname (default 127.0.0.1)
-p Server port (default 6379)
-s Server socket (overrides host and port)
-a Password for Redis Auth
-c Number of parallel connections (default 50)
-n Total number of requests (default 100000)
-d Data size of SET/GET value in bytes (default 2)
--dbnum SELECT the specified db number (default 0)
-k 1=keep alive 0=reconnect (default 1)
-r Use random keys for SET/GET/INCR, random values for SADD
Using this option the benchmark will expand the string __rand_int__
inside an argument with a 12 digits number in the specified range
from 0 to keyspacelen-1. The substitution changes every time a command
is executed. Default tests use this to hit random keys in the
specified range.
-P Pipeline requests. Default 1 (no pipeline).
-e If server replies with errors, show them on stdout.
(no more than 1 error per second is displayed)
-q Quiet. Just show query/sec values
--csv Output in CSV format
-l Loop. Run the tests forever
-t Only run the comma separated list of tests. The test
names are the same as the ones produced as output.
-I Idle mode. Just open N idle connections and wait.

Examples:
Run the benchmark with the default configuration against 127.0.0.1:6379:
# 运行默认配置下的测试
$ redis-benchmark
Use 20 parallel clients, for a total of 100k requests, against 192.168.1.1:
# 指定并发数20,总请求数为10W,redis server主机IP为192.168.1.1
$ redis-benchmark -h 192.168.1.1 -p 6379 -n 100000 -c 20
Fill 127.0.0.1:6379 with about 1 million keys only using the SET test:
# 测试SET随机数性能
$ redis-benchmark -t set -n 1000000 -r 100000000
Benchmark 127.0.0.1:6379 for a few commands producing CSV output:
# 测试结果输出到csv
$ redis-benchmark -t ping,set,get -n 100000 --csv
Benchmark a specific command line:
# 执行特定命令下的测试
$ redis-benchmark -r 10000 -n 10000 eval 'return redis.call("ping")' 0
Fill a list with 10000 random elements:
# 测试list入队的速度
$ redis-benchmark -r 10000 -n 10000 lpush mylist __rand_int__
On user specified command lines __rand_int__ is replaced with a random integer
with a range of values selected by the -r option.


• 2 实际测试过程

  • redis-benchmark 默认参数下的测试

[root@redis-test-slave ~ ]$ redis-benchmark
====== PING_INLINE ======
100000 requests completed in 0.83 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
120192.30 requests per second

====== PING_BULK ======
100000 requests completed in 0.85 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
118203.30 requests per second

====== SET ======
100000 requests completed in 0.80 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
125786.16 requests per second

====== GET ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
125944.58 requests per second

====== INCR ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
126903.55 requests per second

====== LPUSH ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
126262.62 requests per second

====== RPUSH ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
126103.41 requests per second

====== LPOP ======
100000 requests completed in 0.80 seconds
50 parallel clients
3 bytes payload
keep alive: 1

99.97% <= 1 milliseconds
100.00% <= 1 milliseconds
125628.14 requests per second

====== RPOP ======
100000 requests completed in 0.80 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
125786.16 requests per second

====== SADD ======
100000 requests completed in 0.80 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
125786.16 requests per second

====== HSET ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
126103.41 requests per second

====== SPOP ======
100000 requests completed in 0.80 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
125628.14 requests per second

====== LPUSH (needed to benchmark LRANGE) ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
126262.62 requests per second

====== LRANGE_100 (first 100 elements) ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
127388.53 requests per second

====== LRANGE_300 (first 300 elements) ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
127388.53 requests per second

====== LRANGE_500 (first 450 elements) ======
100000 requests completed in 0.78 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
127551.02 requests per second

====== LRANGE_600 (first 600 elements) ======
100000 requests completed in 0.79 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
126742.72 requests per second

====== MSET (10 keys) ======
100000 requests completed in 0.77 seconds
50 parallel clients
3 bytes payload
keep alive: 1

100.00% <= 0 milliseconds
129701.68 requests per second


参考

http://www.redis.cn/topics/benchmarks.html