测试软件：sysbench-memory

1、sysbench测试主机在内存中的连续读写速度
2、测试环境，虚拟机4线程、1G内存，PC服务器96线程、512G内存
虚拟机传输16G数据、块大小16K、线程数4

1. root@pm1:/root# sysbench --test=memory --memory-block-size=16K --memory-total-size=16G --num-threads=4 run
   
2. sysbench 0.5: multi-threaded system evaluation benchmark
   
3. 
   
4. Running the test with following options:
   
5. Number of threads: 4
   
6. Random number generator seed is 0 and will be ignored
   
7. 
   
8. 
   
9. Threads
   
10. 
    
11. Operations performed: 1048576 (269059.42 ops/sec)
    
12. 
    
13. 16384.00 MB transferred (4204.05 MB/sec)
    
14. 
    
15. 
    
16. General statistics:
    
17.     total time: 3.8972s
    
18.     total number of events: 1048576
    
19.     total time taken by event execution: 9.0149s
    
20.     response time:
    
21.          min: 0.00ms
    
22.          avg: 0.01ms
    
23.          max: 29.98ms
    
24.          approx. 95 percentile: 0.01ms
    
25. 
    
26. Threads fairness:
    
27.     events (avg/stddev): 262144.0000/2079.32
    
28.     execution time (avg/stddev): 2.2537/0.02

PC服务器传输16G、块大小16K、线程数96

1. [root@ ~]# sysbench --test=memory --memory-block-size=16K --memory-total-size=16G --num-threads=96 run
   
2. sysbench 0.5: multi-threaded system evaluation benchmark
   
3. 
   
4. Running the test with following options:
   
5. Number of threads: 96
   
6. Random number generator seed is 0 and will be ignored
   
7. 
   
8. 
   
9. Threads
   
10. 
    
11. Operations performed: 1048576 (405615.58 ops/sec)
    
12. 
    
13. 16384.00 MB transferred (6337.74 MB/sec)
    
14. 
    
15. 
    
16. General statistics:
    
17.     total time: 2.5851s
    
18.     total number of events: 1048576
    
19.     total time taken by event execution: 47.7844s
    
20.     response time:
    
21.          min: 0.00ms
    
22.          avg: 0.05ms
    
23.          max: 8.60ms
    
24.          approx. 95 percentile: 0.08ms
    
25. 
    
26. Threads fairness:
    
27.     events (avg/stddev): 10922.6667/408.97
    
28.     execution time (avg/stddev): 0.4978/0.02

看来多线程在内存中的调度还是很大优势的，如果把PC服务器换成4线程，其实和虚拟机性能差不多。

1. [root@ ~]# sysbench --test=memory --memory-block-size=16K --memory-total-size=16G --num-threads=4 run
   
2. sysbench 0.5: multi-threaded system evaluation benchmark
   
3. 
   
4. Running the test with following options:
   
5. Number of threads: 4
   
6. Random number generator seed is 0 and will be ignored
   
7. 
   
8. 
   
9. Threads
   
10. 
    
11. Operations performed: 1048576 (271394.71 ops/sec)
    
12. 
    
13. 16384.00 MB transferred (4240.54 MB/sec)
    
14. 
    
15. 
    
16. General statistics:
    
17.     total time: 3.8637s
    
18.     total number of events: 1048576
    
19.     total time taken by event execution: 13.2885s
    
20.     response time:
    
21.          min: 0.00ms
    
22.          avg: 0.01ms
    
23.          max: 0.19ms
    
24.          approx. 95 percentile: 0.04ms
    
25. 
    
26. Threads fairness:
    
27.     events (avg/stddev): 262144.0000/11704.01
    
28.     execution time (avg/stddev): 3.3221/0.03
    
29. 
    
30. You have new mail in /var/spool/mail/root

上面是把PC服务器的线程数改成4之后，每秒传输4240M，跟之前虚拟机的传输差不多，但是在最大响应时间差别比较大。

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/12219480/viewspace-1745306/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/12219480/viewspace-1745306/