飞腾PC机器与奔腾G645的PC机器以及E5-2630V4虚拟机内存性能简单对比

1. 对比工具 sysbench

2. 对比命令

 sysbench --test=memory  --memory-total-size=12G --num-threads= run

3 安装方式

#ubuntu
apt-get install sysbench
#centos
yum install epel-release
#然后
yum install sysbench

4 注意点

飞腾PC机器 认不到memory-block-size 的参数 所以只能用默认参数执行

5.1 飞腾机器

root@lccy-:~# sysbench --test=memory  --memory-total-size=12G --num-threads= run
sysbench 0.4.:  multi-threaded system evaluation benchmark

Running the test with following options:
Number of threads: 

Doing memory operations speed test
Memory block size: 1K

Memory transfer size: 12288M

Memory operations type: write
Memory scope type: global
Threads started!
Done.

Operations performed:  (2580688.35 ops/sec)

12288.00 MB transferred (2520.20 MB/sec)

Test execution summary:
    total time:                          .8758s
    total number of events:
    total time taken by event execution: 27.8384
    per-request statistics:
         min:                                  .00ms
         avg:                                  .00ms
         max:                                 .02ms
         approx.   percentile:               .00ms

Threads fairness:
    events (avg/stddev):           1572864.0000/93904.61
    execution time (avg/stddev):   3.4798/0.14

5.2 奔腾G645 2011年的PC

[root@CentOS75 ~]#  sysbench --test=memory  --memory-total-size=12G --num-threads= run
WARNING: the --test option is deprecated. You can pass a script name or path on the command line without any options.
WARNING: --num-threads is deprecated, use --threads instead
sysbench 1.0. (using system LuaJIT 2.0.)

Running the test with following options:
Number of threads:
Initializing random number generator from current time

Running memory speed test with the following options:
  block size: 1KiB
  total size: 12288MiB
  operation: write
  scope: global

Initializing worker threads...

Threads started!

Total operations:  (271858.64 per second)

2655.38 MiB transferred (265.49 MiB/sec)

General statistics:
    total time:                          .0001s
    total number of events:              

Latency (ms):
         min:                                    0.00
         avg:                                    0.01
         max:                                   29.46
         95th percentile:                        0.00
         sum:                                27632.28

Threads fairness:
    events (avg/stddev):           339889.0000/10693.15
    execution time (avg/stddev):   3.4540/0.12

5.3 虚拟机的速度

sysbench 1.0. (using system LuaJIT 2.0.)

Running the test with following options:
Number of threads:
Initializing random number generator from current time

Running memory speed test with the following options:
  block size: 1KiB
  total size: 12288MiB
  operation: write
  scope: global

Initializing worker threads...

Threads started!

Total operations:  (9964944.89 per second)

12288.00 MiB transferred (9731.39 MiB/sec)

General statistics:
    total time:                          .2607s
    total number of events:              

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    0.64
         95th percentile:                        0.00
         sum:                                 7497.49

Threads fairness:
    events (avg/stddev):           1572864.0000/0.00
    execution time (avg/stddev):   0.9372/0.00

5.4 某商四路服务器的速度

[root@ /]#  sysbench --test=memory  --memory-total-size=12G --num-threads= run
WARNING: the --test option is deprecated. You can pass a script name or path on the command line without any options.
WARNING: --num-threads is deprecated, use --threads instead
sysbench 1.0. (using system LuaJIT 2.0.)

Running the test with following options:
Number of threads:
Initializing random number generator from current time

Running memory speed test with the following options:
  block size: 1KiB
  total size: 12288MiB
  operation: write
  scope: global

Initializing worker threads...

Threads started!

Total operations:  (1530058.55 per second)

12288.00 MiB transferred (1494.20 MiB/sec)

General statistics:
    total time:                          .2184s
    total number of events:              

Latency (ms):
         min:                                    0.00
         avg:                                    0.00
         max:                                    1.02
         95th percentile:                        0.01
         sum:                                61058.16

Threads fairness:
    events (avg/stddev):           1572864.0000/0.00
    execution time (avg/stddev):   7.6323/0.07

我理解不了 为什么四路服务器的性能这么差。。。

6. 龙芯机器的安装与测试:

github 上面下载压缩包
https://github.com/akopytov/sysbench

解压缩
然后进行处理

./autogen.sh

然后
./configure --without-mysql

执行
make

执行失败

make install
Detecting operating system.......success [linux]
Detecting machine architecture...failed [unsupported]
make[1]: *** [lib/libck.a] 错误 1
make[1]: 离开目录“/root/sysbench-1.0.17/third_party/concurrency_kit”
make: *** [all-recursive] 错误 1
[root@NeoKylin sysbench-1.0.17]# ll