MySQL优化系列4-MySQL压力测试

爱被打了一巴掌 2022-10-10 12:46 278阅读 0赞

备注:测试数据库版本为MySQL 8.0

### 文章目录 ###

*  一.MySQL压力测试概述
 *  二.压力测试的工具
 *   *  2.1 测试工具-fio
     *   *  2.1.1 FIO安装
         *  2.1.2 fio参数说明
         *  2.1.3 测试用例
     *  2.2 TPCC测试
     *   *  2.2.1 环境安装
         *  2.2.2 bug修复
         *  2.2.3 开始测试
     *  2.3 Sysbench
     *   *  2.3.1 sysbench安装
         *  2.3.2 sysbench参数
         *  2.3.3 测试CPU
         *  2.3.4 测试IO
         *  2.3.5 对mysql事务型OLTP的测试
 *  参考:

# 一.MySQL压力测试概述 #

为什么压力测试很重要？因为压力测试是唯一方便有效的、可以学习系统在给定的工作负载下会发生什么的方法。压力测试可以观察系统在不同压力下的行为，评估系统的容量，掌握哪些是重要的变化，或者观察系统如何处理不同的数据。基准测试可以在系统实际负载之外创造一些虚构场景进行测试。

那么，我们测试一般测试哪些指标呢?

1.  吞吐量  
    吞吐量指的是单位时间内的事务处理数，常用的测试单位是每秒事务数（TPS）。
2.  响应时间或者延迟  
    这个指标用于测试任务所需的整体时间。
3.  并发性  
    并发性是一个非常重要又经常被误解和误用的指标。  
    一个Web站点“同时有50000个用户”访问，却可能只有10～15个并发请求到MySQL数据库，因为每个用户并不是时时刻刻的都在发送请求到数据库。
4.  可扩展性  
    在系统的业务压力可能发生变化的情况下，测试可扩展性就非常必要了。

# 二.压力测试的工具 #

## 2.1 测试工具-fio ##

FIO是测试IOPS的非常好的工具，用来对硬件进行压力测试和验证，支持13种不同的I/O 引擎，包括:sync,mmap, libaio, posixaio, SG v3, splice, null, network, syslet, guasi, solarisaio 等等。  
fio 官网地址：[http://freecode.com/projects/fio][http_freecode.com_projects_fio]

### 2.1.1 FIO安装 ###

wget http://brick.kernel.dk/snaps/fio-3.18.tar.gz
    yum install libaio-devel
    tar -zxvf fio-3.18.tar.gz
    cd fio-3.18
    make
    make install

### 2.1.2 fio参数说明 ###

<table> 
 <thead> 
  <tr> 
   <th>参数</th> 
   <th>参数说明</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>filename=/dev/mapper/centos-home</td> 
   <td>测试文件名称，通常选择需要测试的盘的data目录</td> 
  </tr> 
  <tr> 
   <td>direct=1</td> 
   <td>测试过程绕过机器自带的buffer。使测试结果更真实。</td> 
  </tr> 
  <tr> 
   <td>rw</td> 
   <td>rw=randwread 测试随机读的I/O<br>rw=randwrite 测试随机写的I/O<br>rw=randrw 测试随机混合写和读的I/O<br>rw=read 测试顺序读的I/O<br>rw=write 测试顺序写的I/O<br>rw=rw 测试顺序混合写和读的I/O</td> 
  </tr> 
  <tr> 
   <td>bs=16k</td> 
   <td>单次io的块文件大小为16k</td> 
  </tr> 
  <tr> 
   <td>bsrange=512-2048</td> 
   <td>同上，提定数据块的大小范围</td> 
  </tr> 
  <tr> 
   <td>size=5g</td> 
   <td>本次的测试文件大小为5g，以每次4k的io进行测试。</td> 
  </tr> 
  <tr> 
   <td>numjobs=30</td> 
   <td>本次的测试线程为30.</td> 
  </tr> 
  <tr> 
   <td>runtime=1000</td> 
   <td>测试时间为1000秒，如果不写则一直将5g文件分4k每次写完为止。</td> 
  </tr> 
  <tr> 
   <td>ioengine=psync</td> 
   <td>io引擎使用pync方式</td> 
  </tr> 
  <tr> 
   <td>rwmixwrite=30</td> 
   <td>在混合读写的模式下，写占30%</td> 
  </tr> 
  <tr> 
   <td>group_reporting</td> 
   <td>关于显示结果的，汇总每个进程的信息。</td> 
  </tr> 
  <tr> 
   <td>lockmem=1g</td> 
   <td>只使用1g内存进行测试。</td> 
  </tr> 
  <tr> 
   <td>zero_buffers</td> 
   <td>用0初始化系统buffer。</td> 
  </tr> 
  <tr> 
   <td>nrfiles=8</td> 
   <td>每个进程生成文件的数量</td> 
  </tr> 
 </tbody> 
</table>

### 2.1.3 测试用例 ###

我们来测试一个随机读

代码:

fio -filename=/dev/mapper/centos-home -direct=1 -iodepth 1 -thread -rw=randread -ioengine=psync -bs=16k -size=10G -numjobs=10 -runtime=1000 -group_reporting -name=mytest

测试记录:

[root@10-31-1-119 software]# fio -filename=/dev/mapper/centos-home -direct=1 -iodepth 1 -thread -rw=randread -ioengine=psync -bs=16k -size=10G -numjobs=10 -runtime=1000 -group_reporting -name=mytest
    mytest: (g=0): rw=randread, bs=(R) 16.0KiB-16.0KiB, (W) 16.0KiB-16.0KiB, (T) 16.0KiB-16.0KiB, ioengine=psync, iodepth=1
    ...
    fio-3.18
    Starting 10 threads
    Jobs: 1 (f=1): [_(9),r(1)][100.0%][r=637MiB/s][r=40.8k IOPS][eta 00m:00s]               
    mytest: (groupid=0, jobs=10): err= 0: pid=5178: Wed May 12 11:01:45 2021
      read: IOPS=67.4k, BW=1053MiB/s (1104MB/s)(100GiB/97218msec)
        clat (usec): min=16, max=12159, avg=144.90, stdev=161.22
         lat (usec): min=17, max=12159, avg=145.05, stdev=161.23
        clat percentiles (usec):
         |  1.00th=[   38],  5.00th=[   52], 10.00th=[   62], 20.00th=[   76],
         | 30.00th=[   84], 40.00th=[   90], 50.00th=[   96], 60.00th=[  103],
         | 70.00th=[  112], 80.00th=[  124], 90.00th=[  400], 95.00th=[  545],
         | 99.00th=[  742], 99.50th=[  857], 99.90th=[ 1237], 99.95th=[ 1352],
         | 99.99th=[ 2147]
       bw (  MiB/s): min=  746, max= 1398, per=100.00%, avg=1067.44, stdev= 9.28, samples=1908
       iops        : min=47780, max=89506, avg=68316.38, stdev=594.14, samples=1908
      lat (usec)   : 20=0.01%, 50=4.46%, 100=51.66%, 250=32.91%, 500=4.35%
      lat (usec)   : 750=5.67%, 1000=0.64%
      lat (msec)   : 2=0.29%, 4=0.01%, 10=0.01%, 20=0.01%
      cpu          : usr=2.07%, sys=14.06%, ctx=6539035, majf=0, minf=57
      IO depths    : 1=100.0%, 2=0.0%, 4=0.0%, 8=0.0%, 16=0.0%, 32=0.0%, >=64=0.0%
         submit    : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
         complete  : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
         issued rwts: total=6553600,0,0,0 short=0,0,0,0 dropped=0,0,0,0
         latency   : target=0, window=0, percentile=100.00%, depth=1
    
    Run status group 0 (all jobs):
       READ: bw=1053MiB/s (1104MB/s), 1053MiB/s-1053MiB/s (1104MB/s-1104MB/s), io=100GiB (107GB), run=97218-97218msec
    [root@10-31-1-119 software]#

## 2.2 TPCC测试 ##

TPC(Tracsaction Processing Performance Council) 事务处理性能协会是一个评价大型数据库系统软硬件性能的非盈利的组织,TPC-C是TPC协会制定的，用来测试典型的复杂OLTP系统的性能。Tpcc-mysql是percona基于tpcc衍生出来的产品，专用于mysql基准测试.

### 2.2.1 环境安装 ###

下载源码:

-- 如果失败，多试几次即可
    git clone https://github.com/Percona-Lab/tpcc-mysql.git

下载记录:

[root@10-31-1-120 software]# git clone https://github.com/Percona-Lab/tpcc-mysql.git
    正克隆到 'tpcc-mysql'...
    remote: Enumerating objects: 392, done.
    remote: Total 392 (delta 0), reused 0 (delta 0), pack-reused 392
    接收对象中: 100% (392/392), 202.81 KiB | 9.00 KiB/s, done.
    处理 delta 中: 100% (216/216), done.
    [root@10-31-1-120 software]#

编译：

yum -y install mysql-devel
    cd tpcc-mysql/src
    make

加载数据：

mysqladmin -u root -p create tpcc1000
    
    mysql -uroot -p
    use tpcc1000;
    source create_table.sql
    source add_fkey_idx.sql
    
    -- 20分钟
    ./tpcc_load -h127.0.0.1 -d tpcc1000 -u root -p passwd -w 100

### 2.2.2 bug修复 ###

运行tpcc\_start的时候会报错:  
2002, HY000. Can’t connect to local MySQL server through socket Vvar. run/mysqld/mysqld.sock’

tpcc-mysql存在2个bug,需要手动修改：  
**文件1： main.c**

char db_socket[DB_STRING_MAX] = " ";
    -- snip -- 
    　　 if(is_local==1){
    
    　　 / exec sql connect :connect_string; /
    
    　　resp = mysql_real_connect(ctx[t_num], "localhost", db_user, db_password, db_string_full, port, db_socket, 0);
    
    }else{
    
    　　/ exec sql connect :connect_string USING :db_string; /
    
    　　resp = mysql_real_connect(ctx[t_num], connect_string, db_user, db_password, db_string_full, port, db_socket, 0);
    
    }

修改为:

char db_socket[DB_STRING_MAX] = "/var/lib/mysql/mysql.sock";
    
    --snip--
    
    　　 if(is_local==1){
    
    　　 / exec sql connect :connect_string; /
    
    　　mysql_thread_init();
    
    　　resp = mysql_real_connect(ctx[t_num], "localhost", db_user, db_password, db_string_full, port, db_socket, 0);
    
    }else{
    
    　　/ exec sql connect :connect_string USING :db_string; /
    
    　　mysql_thread_init();
    
    　　resp = mysql_real_connect(ctx[t_num], connect_string, db_user, db_password, db_string_full, port, db_socket, 0);
    
    }

**文件二：load.c**  
源文件:

if(is_local==1){
    
            /* exec sql connect :connect_string; */
    
            resp = mysql_real_connect(mysql, "localhost", db_user, db_password, db_string, port, NULL, 0);
    
        }else{
    
            /* exec sql connect :connect_string USING :db_string; */
    
            resp = mysql_real_connect(mysql, connect_string, db_user, db_password, db_string, port, NULL, 0);

修改为:

if(is_local==1){
                /* exec sql connect :connect_string; */
                mysql_thread_init();
                resp = mysql_real_connect(mysql, "localhost", db_user, db_password, db_string, port, "/var/lib/mysql/mysql.sock", 0);
            }else{
                mysql_thread_init();
                /* exec sql connect :connect_string USING :db_string; */
                resp = mysql_real_connect(mysql, connect_string, db_user, db_password, db_string, port, "/var/lib/mysql/mysql.sock", 0);
            }

### 2.2.3 开始测试 ###

参数含义  
　　　　-h server\_host: 服务器名  
　　　　-P port : 端口号，默认为3306  
　　　　-d database\_name: 数据库名  
　　　　-u mysql\_user : 用户名  
　　　　-p mysql\_password : 密码  
　　　　-w warehouses: 仓库的数量  
　　　　-c connections : 线程数，默认为1  
　　　　-r warmup\_time : 热身时间，单位:s，默认为10s ,热身是为了将数据加载到内存。  
　　　　-l running\_time: 测试时间，单位:s，默认为20s  
　　　　-i report\_interval: 指定生成报告间隔时长  
　　　　-f report\_file: 测试结果输出文件

代码:

./tpcc_start -h 10.31.1.120 -d tpcc1000 -u root -p "abc123"  -w 10 -c 5 -r 120 -l 600 -i 10  -f tpcc_mysql_20210524.log

报告解读:

[root@10-31-1-120 tpcc-mysql]# ./tpcc_start -h 10.31.1.120 -d tpcc1000 -u root -p "abc123"  -w 10 -c 5 -r 120 -l 600 -i 10  -f tpcc_mysql_20210524.log    
    ***************************************
    *** ###easy### TPC-C Load Generator ***
    ***************************************
    option h with value '10.31.1.120'                       
    option d with value 'tpcc1000'                          
    option u with value 'root'                              
    option p with value 'abc123'                            
    option w with value '10'                                
    option c with value '5'                                 
    option r with value '120'                               
    option l with value '600'                               
    option i with value '10'                                
    option f with value 'tpcc_mysql_20210524.log'
    <Parameters>
         [server]: 10.31.1.120                              -- 主机
         [port]: 3306                                       -- 端口
         [DBname]: tpcc1000                                 -- 数据库
           [user]: root                                     -- 账号
           [pass]: abc123                                   -- 密码
      [warehouse]: 10                                       -- 仓库数
     [connection]: 5                                        -- 并发线程数 
         [rampup]: 120 (sec.)                               -- 数据预热时长 
        [measure]: 600 (sec.)                               -- 压测时长
    
    RAMP-UP TIME.(120 sec.)                     --预热结束
    
    MEASURING START.                            --开始压测
    
      10, trx: 1585, 95%: 40.465, 99%: 68.161, max_rt: 154.953, 1585|108.418, 158|68.316, 158|180.866, 159|290.456
      20, trx: 1631, 95%: 42.020, 99%: 66.290, max_rt: 134.796, 1635|90.688, 164|19.419, 163|166.759, 163|103.197
      30, trx: 1685, 95%: 42.539, 99%: 63.760, max_rt: 95.661, 1682|66.667, 168|23.617, 170|124.743, 169|78.784
      40, trx: 1717, 95%: 41.334, 99%: 61.676, max_rt: 101.083, 1718|75.835, 172|17.605, 170|153.279, 171|110.645
      50, trx: 1697, 95%: 40.332, 99%: 65.344, max_rt: 113.429, 1695|107.809, 170|18.856, 170|191.108, 169|117.728
      60, trx: 1334, 95%: 52.189, 99%: 93.558, max_rt: 1590.916, 1337|1414.192, 132|13.973, 134|1496.522, 134|113.481
      70, trx: 1568, 95%: 44.187, 99%: 64.200, max_rt: 105.029, 1567|89.779, 157|19.481, 157|197.804, 157|129.003
      80, trx: 1592, 95%: 44.200, 99%: 77.131, max_rt: 133.690, 1594|88.494, 159|15.762, 159|155.874, 160|117.515
      90, trx: 1662, 95%: 41.038, 99%: 69.584, max_rt: 113.852, 1656|88.420, 167|11.431, 166|152.061, 165|117.842
     100, trx: 1667, 95%: 41.907, 99%: 64.028, max_rt: 136.287, 1672|91.327, 167|11.070, 167|173.430, 167|116.591
     110, trx: 1681, 95%: 41.284, 99%: 69.542, max_rt: 132.606, 1683|68.603, 168|55.347, 168|162.181, 168|116.353
     120, trx: 1549, 95%: 45.625, 99%: 73.700, max_rt: 135.726, 1542|85.992, 155|15.106, 154|219.304, 155|83.339
     130, trx: 1574, 95%: 45.177, 99%: 66.968, max_rt: 158.751, 1578|54.682, 157|13.111, 159|170.681, 157|78.849
     140, trx: 1546, 95%: 44.854, 99%: 83.724, max_rt: 145.061, 1544|77.227, 155|22.305, 154|181.884, 155|139.306
     150, trx: 1635, 95%: 44.639, 99%: 74.343, max_rt: 165.761, 1637|82.254, 163|43.885, 163|186.883, 163|116.960
     160, trx: 1642, 95%: 42.488, 99%: 65.344, max_rt: 108.625, 1644|100.719, 165|12.851, 165|180.322, 165|148.575
     170, trx: 1643, 95%: 44.572, 99%: 69.688, max_rt: 118.454, 1636|82.954, 163|16.305, 164|151.805, 164|79.632
     180, trx: 1544, 95%: 46.871, 99%: 83.549, max_rt: 143.344, 1551|71.973, 155|19.426, 154|166.200, 155|114.349
     190, trx: 1529, 95%: 51.122, 99%: 76.327, max_rt: 119.329, 1528|90.759, 153|37.335, 153|202.649, 153|119.973
     200, trx: 1516, 95%: 45.913, 99%: 76.190, max_rt: 129.581, 1518|103.034, 152|15.985, 152|186.119, 151|110.922
     210, trx: 1629, 95%: 43.414, 99%: 62.907, max_rt: 116.559, 1622|91.839, 162|21.515, 163|204.726, 163|132.465
     220, trx: 1661, 95%: 43.284, 99%: 66.409, max_rt: 117.584, 1659|125.108, 166|22.964, 166|190.968, 166|101.227
     230, trx: 1639, 95%: 41.186, 99%: 68.039, max_rt: 139.412, 1642|81.764, 165|19.266, 164|201.031, 164|124.726
     240, trx: 1529, 95%: 47.110, 99%: 74.254, max_rt: 127.122, 1528|89.788, 152|16.630, 153|185.137, 153|130.613
     250, trx: 1525, 95%: 51.122, 99%: 81.109, max_rt: 116.085, 1527|78.839, 153|20.238, 152|224.054, 152|104.700
     260, trx: 1541, 95%: 50.817, 99%: 83.724, max_rt: 130.356, 1535|87.927, 153|24.012, 154|201.159, 154|84.816
     270, trx: 1383, 95%: 41.657, 99%: 80.528, max_rt: 1485.578, 1388|1473.847, 139|19.837, 138|1589.119, 138|122.915
     280, trx: 1630, 95%: 44.333, 99%: 69.376, max_rt: 97.787, 1630|80.551, 163|12.524, 163|177.245, 164|118.633
     290, trx: 1682, 95%: 40.915, 99%: 65.776, max_rt: 119.428, 1681|78.685, 168|23.446, 168|160.553, 169|115.521
     300, trx: 1509, 95%: 48.600, 99%: 77.664, max_rt: 104.815, 1513|89.817, 152|23.437, 152|160.403, 150|120.316
     310, trx: 1448, 95%: 51.091, 99%: 81.890, max_rt: 133.332, 1448|95.003, 144|23.381, 145|184.240, 145|111.933
     320, trx: 1551, 95%: 44.453, 99%: 75.013, max_rt: 177.458, 1552|83.327, 155|71.883, 154|176.311, 155|103.884
     330, trx: 1595, 95%: 46.424, 99%: 74.767, max_rt: 101.033, 1590|100.878, 160|19.500, 160|161.520, 160|116.336
     340, trx: 1620, 95%: 44.519, 99%: 69.293, max_rt: 117.582, 1622|82.830, 161|25.185, 162|145.765, 161|147.591
     350, trx: 1631, 95%: 40.610, 99%: 72.671, max_rt: 110.570, 1631|99.961, 164|13.384, 163|182.350, 164|91.907
     360, trx: 1512, 95%: 47.124, 99%: 74.032, max_rt: 112.626, 1512|77.888, 151|18.268, 151|193.651, 151|122.528
     370, trx: 1495, 95%: 48.556, 99%: 69.730, max_rt: 97.009, 1494|92.760, 150|23.593, 149|170.044, 149|83.201
     380, trx: 1529, 95%: 46.675, 99%: 71.356, max_rt: 101.201, 1532|67.657, 152|15.640, 154|172.814, 154|87.031
     390, trx: 1635, 95%: 41.223, 99%: 72.107, max_rt: 151.944, 1629|87.851, 164|35.737, 163|169.300, 162|132.981
     400, trx: 1632, 95%: 44.055, 99%: 70.317, max_rt: 92.558, 1637|95.022, 163|43.889, 164|210.165, 164|110.684
     410, trx: 1652, 95%: 43.492, 99%: 70.338, max_rt: 126.571, 1650|94.306, 165|16.432, 164|163.375, 165|122.867
     420, trx: 1560, 95%: 46.913, 99%: 81.133, max_rt: 142.466, 1560|105.641, 156|17.668, 157|199.629, 156|135.119
     430, trx: 1536, 95%: 47.749, 99%: 83.699, max_rt: 136.907, 1537|81.965, 154|28.370, 153|179.660, 154|143.256
     440, trx: 1563, 95%: 46.244, 99%: 75.509, max_rt: 126.649, 1564|109.099, 156|15.467, 157|171.503, 155|142.518
     450, trx: 1600, 95%: 45.913, 99%: 71.978, max_rt: 100.656, 1599|73.022, 160|36.008, 160|191.457, 161|82.701
     460, trx: 1635, 95%: 44.399, 99%: 82.283, max_rt: 117.744, 1633|79.148, 163|19.903, 162|188.135, 163|102.372
     470, trx: 1676, 95%: 44.174, 99%: 68.962, max_rt: 108.584, 1675|69.334, 168|18.741, 168|159.369, 167|74.830
     480, trx: 1552, 95%: 45.339, 99%: 69.501, max_rt: 114.857, 1557|90.775, 156|38.205, 156|236.356, 156|85.198
     490, trx: 1519, 95%: 48.252, 99%: 82.628, max_rt: 204.636, 1515|139.115, 151|22.979, 151|194.910, 151|101.544
     500, trx: 1553, 95%: 49.083, 99%: 69.148, max_rt: 134.515, 1556|87.782, 156|28.435, 156|189.380, 156|132.119
     510, trx: 1634, 95%: 42.412, 99%: 68.735, max_rt: 115.450, 1635|74.971, 163|12.749, 163|187.160, 164|123.722
     520, trx: 1631, 95%: 43.323, 99%: 73.414, max_rt: 108.546, 1627|90.988, 163|17.812, 164|168.162, 163|124.428
     530, trx: 1600, 95%: 44.693, 99%: 71.164, max_rt: 105.214, 1601|74.805, 160|21.664, 159|161.137, 159|133.575
     540, trx: 1458, 95%: 49.644, 99%: 78.177, max_rt: 124.618, 1460|96.897, 147|30.464, 147|180.685, 147|121.754
     550, trx: 1511, 95%: 46.065, 99%: 72.518, max_rt: 130.639, 1507|82.385, 150|19.416, 150|241.390, 151|106.991
     560, trx: 1561, 95%: 44.679, 99%: 74.143, max_rt: 116.987, 1564|85.479, 156|14.463, 157|239.011, 156|74.772
     570, trx: 1593, 95%: 42.603, 99%: 68.714, max_rt: 121.134, 1592|127.673, 159|21.745, 158|182.691, 159|96.393
     580, trx: 1591, 95%: 44.719, 99%: 79.285, max_rt: 117.820, 1591|85.518, 159|75.231, 160|186.695, 159|98.639
     590, trx: 1588, 95%: 44.827, 99%: 71.356, max_rt: 123.872, 1586|115.386, 159|15.790, 158|177.794, 159|112.341
     600, trx: 1232, 95%: 48.644, 99%: 79.952, max_rt: 1518.349, 1234|115.769, 124|16.189, 123|209.123, 123|122.170
    
    
    STOPPING THREADS.....
    
    <Raw Results>
      [0] sc:2957 lt:91661  rt:0  fl:0 avg_rt: 19.5 (5)
      [1] sc:62192 lt:32425  rt:0  fl:0 avg_rt: 7.6 (5)
      [2] sc:8800 lt:662  rt:0  fl:0 avg_rt: 2.2 (5)
      [3] sc:7371 lt:2090  rt:0  fl:0 avg_rt: 67.0 (80)
      [4] sc:1797 lt:7665  rt:0  fl:0 avg_rt: 40.3 (20)
     in 600 sec.
    
    <Raw Results2(sum ver.)>
      [0] sc:2957  lt:91661  rt:0  fl:0 
      [1] sc:62194  lt:32425  rt:0  fl:0 
      [2] sc:8800  lt:662  rt:0  fl:0 
      [3] sc:7371  lt:2090  rt:0  fl:0 
      [4] sc:1797  lt:7665  rt:0  fl:0 
    
    <Constraint Check> (all must be [OK])
     [transaction percentage]
            Payment: 43.48% (>=43.0%) [OK]
       Order-Status: 4.35% (>= 4.0%) [OK]
           Delivery: 4.35% (>= 4.0%) [OK]
        Stock-Level: 4.35% (>= 4.0%) [OK]
     [response time (at least 90% passed)]
          New-Order: 3.13%  [NG] *
            Payment: 65.73%  [NG] *
       Order-Status: 93.00%  [OK]
           Delivery: 77.91%  [NG] *
        Stock-Level: 18.99%  [NG] *
    
    <TpmC>
                     9461.800 TpmC

10, trx: 1585, 95%: 40.465, 99%: 68.161, max\_rt: 154.953, 1585|108.418, 158|68.316, 158|180.866, 159|290.456

逗号进行分割。

10 表示从测试开始到现在的时间为10秒。  
trx: 1585 表示这个时间间隔完成的新订单事物数量，实际上就是吞吐量，越大越好。  
95%: 40.465 表示这个时间间隔内95%的新订单事物响应时间为40.465秒。  
99%: 68.161 表示这个时间间隔内99%的新订单事物响应时间为68.161秒。  
max\_rt: 154.953 表示这个时间间隔内新订单事物最大响应时间为154.953秒。  
1585|108.418, 158|68.316, 158|180.866, 159|290.456 这几个表示其他种类事物的吞吐量和响应时间，可以忽略。

\[0\] sc:2957 lt:91661 rt:0 fl:0 avg\_rt: 19.5 (5) //表示新订单业务成功(success,简写sc)次数，延迟(late,简写lt)次数，重试(retry,简写rt)次数，失败(failure,简写fl)次数。  
\[1\] sc:62192 lt:32425 rt:0 fl:0 avg\_rt: 7.6 (5) //表示支付业务成功(success,简写sc)次数，延迟(late,简写lt)次数，重试(retry,简写rt)次数，失败(failure,简写fl)次数。  
\[2\] sc:8800 lt:662 rt:0 fl:0 avg\_rt: 2.2 (5) //表示订单状态业务成功(success,简写sc)次数，延迟(late,简写lt)次数，重试(retry,简写rt)次数，失败(failure,简写fl)次数。  
\[3\] sc:7371 lt:2090 rt:0 fl:0 avg\_rt: 67.0 (80) //表示发货业务成功(success,简写sc)次数，延迟(late,简写lt)次数，重试(retry,简写rt)次数，失败(failure,简写fl)次数。  
\[4\] sc:1797 lt:7665 rt:0 fl:0 avg\_rt: 40.3 (20) //表示库存业务成功(success,简写sc)次数，延迟(late,简写lt)次数，重试(retry,简写rt)次数，失败(failure,简写fl)次数。

(all must be \[OK\]) //下面所有的结果必须为OK  
\[transaction percentage\]  
Payment: 43.48% (>=43.0%) \[OK\] //支付模块占总业务的43.48%  
Order-Status: 4.35% (>= 4.0%) \[OK\] //订单状态模块占总业务的4.35%  
Delivery: 4.35% (>= 4.0%) \[OK\] //发货模块占总业务的4.35%  
Stock-Level: 4.35% (>= 4.0%) \[OK\] //库存模块占总业务的4.35%  
\[response time (at least 90% passed)\] //响应时间是否满足  
New-Order: 3.13% \[NG\] \*  
Payment: 65.73% \[NG\] \*  
Order-Status: 93.00% \[OK\]  
Delivery: 77.91% \[NG\] \*  
Stock-Level: 18.99% \[NG\] \*

指标如下：  
New-Order 5ms  
Payment 5ms  
Order-Status 5ms  
Delivery 80ms  
Stock-Level 20ms

9461.800 TpmC //每分钟能够处理的新订单数量。

## 2.3 Sysbench ##

sysbench可以执行多种类型的基准测试，它不仅设计用来测试数据库的性能，也可以测试运行数据库的服务器的性能。

### 2.3.1 sysbench安装 ###

一般推荐使用yum进行安装

yum -y install sysbench

### 2.3.2 sysbench参数 ###

直接查看帮助即可看到sysbench支持的参数及含义

[root@10-31-1-120 ~]# sysbench --help
    Usage:
      sysbench [options]... [testname] [command]
    
    Commands implemented by most tests: prepare run cleanup help
    
    General options:
      --threads=N                     number of threads to use [1]
      --events=N                      limit for total number of events [0]
      --time=N                        limit for total execution time in seconds [10]
      --forced-shutdown=STRING        number of seconds to wait after the --time limit before forcing shutdown, or 'off' to disable [off]
      --thread-stack-size=SIZE        size of stack per thread [64K]
      --rate=N                        average transactions rate. 0 for unlimited rate [0]
      --report-interval=N             periodically report intermediate statistics with a specified interval in seconds. 0 disables intermediate reports [0]
      --report-checkpoints=[LIST,...] dump full statistics and reset all counters at specified points in time. The argument is a list of comma-separated values representing the amount of time in seconds elapsed from start of test when report checkpoint(s) must be performed. Report checkpoints are off by default. []
      --debug[=on|off]                print more debugging info [off]
      --validate[=on|off]             perform validation checks where possible [off]
      --help[=on|off]                 print help and exit [off]
      --version[=on|off]              print version and exit [off]
      --config-file=FILENAME          File containing command line options
      --tx-rate=N                     deprecated alias for --rate [0]
      --max-requests=N                deprecated alias for --events [0]
      --max-time=N                    deprecated alias for --time [0]
      --num-threads=N                 deprecated alias for --threads [1]
    
    Pseudo-Random Numbers Generator options:
      --rand-type=STRING random numbers distribution {uniform,gaussian,special,pareto} [special]
      --rand-spec-iter=N number of iterations used for numbers generation [12]
      --rand-spec-pct=N  percentage of values to be treated as 'special' (for special distribution) [1]
      --rand-spec-res=N  percentage of 'special' values to use (for special distribution) [75]
      --rand-seed=N      seed for random number generator. When 0, the current time is used as a RNG seed. [0]
      --rand-pareto-h=N  parameter h for pareto distribution [0.2]
    
    Log options:
      --verbosity=N verbosity level {5 - debug, 0 - only critical messages} [3]
    
      --percentile=N       percentile to calculate in latency statistics (1-100). Use the special value of 0 to disable percentile calculations [95]
      --histogram[=on|off] print latency histogram in report [off]
    
    General database options:
    
      --db-driver=STRING  specifies database driver to use ('help' to get list of available drivers) [mysql]
      --db-ps-mode=STRING prepared statements usage mode {auto, disable} [auto]
      --db-debug[=on|off] print database-specific debug information [off]
    
    
    Compiled-in database drivers:
      mysql - MySQL driver
      pgsql - PostgreSQL driver
    
    mysql options:
      --mysql-host=[LIST,...]          MySQL server host [localhost]
      --mysql-port=[LIST,...]          MySQL server port [3306]
      --mysql-socket=[LIST,...]        MySQL socket
      --mysql-user=STRING              MySQL user [sbtest]
      --mysql-password=STRING          MySQL password []
      --mysql-db=STRING                MySQL database name [sbtest]
      --mysql-ssl[=on|off]             use SSL connections, if available in the client library [off]
      --mysql-ssl-cipher=STRING        use specific cipher for SSL connections []
      --mysql-compression[=on|off]     use compression, if available in the client library [off]
      --mysql-debug[=on|off]           trace all client library calls [off]
      --mysql-ignore-errors=[LIST,...] list of errors to ignore, or "all" [1213,1020,1205]
      --mysql-dry-run[=on|off]         Dry run, pretend that all MySQL client API calls are successful without executing them [off]
    
    pgsql options:
      --pgsql-host=STRING     PostgreSQL server host [localhost]
      --pgsql-port=N          PostgreSQL server port [5432]
      --pgsql-user=STRING     PostgreSQL user [sbtest]
      --pgsql-password=STRING PostgreSQL password []
      --pgsql-db=STRING       PostgreSQL database name [sbtest]
    
    Compiled-in tests:
      fileio - File I/O test
      cpu - CPU performance test
      memory - Memory functions speed test
      threads - Threads subsystem performance test
      mutex - Mutex performance test
    
    See 'sysbench <testname> help' for a list of options for each test.
    
    [root@10-31-1-120 ~]#

### 2.3.3 测试CPU ###

命令:

sysbench --test=cpu --cpu-max-prime=20000 run

测试记录:

[root@10-31-1-120 ~]# sysbench --test=cpu --cpu-max-prime=20000 run
    WARNING: the --test option is deprecated. You can pass a script name or path on the command line without any options.
    sysbench 1.0.20 (using bundled LuaJIT 2.1.0-beta2)
    
    Running the test with following options:
    Number of threads: 1
    Initializing random number generator from current time
    
    
    Prime numbers limit: 20000
    
    Initializing worker threads...
    
    Threads started!
    
    CPU speed:
        events per second:   398.45
    
    General statistics:
        total time:                          10.0011s
        total number of events:              3987
    
    Latency (ms):
             min:                                    2.37
             avg:                                    2.51
             max:                                    7.70
             95th percentile:                        2.66
             sum:                                 9999.05
    
    Threads fairness:
        events (avg/stddev):           3987.0000/0.00
        execution time (avg/stddev):   9.9990/0.00

### 2.3.4 测试IO ###

文件I/O（fileio）基准测试可以测试系统在不同I/O负载下的性能。这对于比较不同的硬盘驱动器、不同的RAID卡、不同的RAID模式，都很有帮助。可以根据测试结果来调整I/O子系统。文件I/O基准测试模拟了很多InnoDB的I/O特性。

seqwr – 顺序写入  
seqrewr – 顺序重写  
seqrd – 顺序读取  
rndrd – 随机读取  
rndwr – 随机写入  
rdnrw – 混合随机读/写

命令:

sysbench --test=fileio --file-num=16 --file-total-size=2G prepare

测试记录:

[root@10-31-1-120 software]# sysbench --test=fileio --file-num=16 --file-total-size=2G prepare
    WARNING: the --test option is deprecated. You can pass a script name or path on the command line without any options.
    sysbench 1.0.20 (using bundled LuaJIT 2.1.0-beta2)
    
    16 files, 131072Kb each, 2048Mb total
    Creating files for the test...
    Extra file open flags: (none)
    Creating file test_file.0
    Creating file test_file.1
    Creating file test_file.2
    Creating file test_file.3
    Creating file test_file.4
    Creating file test_file.5
    Creating file test_file.6
    Creating file test_file.7
    Creating file test_file.8
    Creating file test_file.9
    Creating file test_file.10
    Creating file test_file.11
    Creating file test_file.12
    Creating file test_file.13
    Creating file test_file.14
    Creating file test_file.15
    2147483648 bytes written in 6.96 seconds (294.30 MiB/sec).
    [root@10-31-1-120 software]#

### 2.3.5 对mysql事务型OLTP的测试 ###

对于mysql的OLTP测试，和file一样，同样需要经历prepare,run,cleanup三个阶段。prepare阶段会在数据库中产生一张指定行数的表，默认表在sbtest架构下，表名为sbtest(sysbench默认生成表的存储引擎为innodb),如创建一张8000万条记录的表：

命令:

CREATE DATABASE `sbtest` DEFAULT CHARACTER SET utf8;
    
    sysbench --oltp-tables-count=5 --oltp_table_size=10000 --mysql-host=localhost --mysql-user=root --mysql-password=abc123 /usr/share/sysbench/tests/include/oltp_legacy/insert.lua prepare

测试记录:

[root@10-31-1-120 ~]# sysbench --oltp-tables-count=5 --oltp_table_size=10000 --mysql-host=localhost --mysql-user=root --mysql-password=abc123 /usr/share/sysbench/tests/include/oltp_legacy/insert.lua prepare
    sysbench 1.0.20 (using bundled LuaJIT 2.1.0-beta2)
    
    Creating table 'sbtest1'...
    Inserting 10000 records into 'sbtest1'
    Creating secondary indexes on 'sbtest1'...
    Creating table 'sbtest2'...
    Inserting 10000 records into 'sbtest2'
    Creating secondary indexes on 'sbtest2'...
    Creating table 'sbtest3'...
    Inserting 10000 records into 'sbtest3'
    Creating secondary indexes on 'sbtest3'...
    Creating table 'sbtest4'...
    Inserting 10000 records into 'sbtest4'
    Creating secondary indexes on 'sbtest4'...
    Creating table 'sbtest5'...
    Inserting 10000 records into 'sbtest5'
    Creating secondary indexes on 'sbtest5'...
    [root@10-31-1-120 ~]# 
    
    mysql> select count(*) from sbtest1;
    +----------+
    | count(*) |
    +----------+
    |    10000 |
    +----------+
    1 row in set (0.00 sec)
    
    mysql> desc sbtest1;
    +-------+------------------+------+-----+---------+----------------+
    | Field | Type             | Null | Key | Default | Extra          |
    +-------+------------------+------+-----+---------+----------------+
    | id    | int(10) unsigned | NO   | PRI | NULL    | auto_increment |
    | k     | int(10) unsigned | NO   | MUL | 0       |                |
    | c     | char(120)        | NO   |     |         |                |
    | pad   | char(60)         | NO   |     |         |                |
    +-------+------------------+------+-----+---------+----------------+
    4 rows in set (0.00 sec)
    
    mysql> 
    mysql> select * from sbtest1 limit 10;  
    +----+------+-------------------------------------------------------------------------------------------------------------------------+-------------------------------------------------------------+
    | id | k    | c                                                                                                                       | pad                                                         |
    +----+------+-------------------------------------------------------------------------------------------------------------------------+-------------------------------------------------------------+
    |  1 | 4993 | 83868641912-28773972837-60736120486-75162659906-27563526494-20381887404-41576422241-93426793964-56405065102-33518432330 | 67847967377-48000963322-62604785301-91415491898-96926520291 |
    |  2 | 5020 | 38014276128-25250245652-62722561801-27818678124-24890218270-18312424692-92565570600-36243745486-21199862476-38576014630 | 23183251411-36241541236-31706421314-92007079971-60663066966 |
    |  3 | 5044 | 33973744704-80540844748-72700647445-87330233173-87249600839-07301471459-22846777364-58808996678-64607045326-48799346817 | 38615512647-91458489257-90681424432-95014675832-60408598704 |
    |  4 | 5021 | 37002370280-58842166667-00026392672-77506866252-09658311935-56926959306-83464667271-94685475868-28264244556-14550208498 | 63947013338-98809887124-59806726763-79831528812-45582457048 |
    |  5 | 4999 | 44257470806-17967007152-32809666989-26174672567-29883439075-95767161284-94957565003-35708767253-53935174705-16168070783 | 34551750492-67990399350-81179284955-79299808058-21257255869 |
    |  6 | 5006 | 37216201353-39109531021-11197415756-87798784755-02463049870-83329763120-57551308766-61100580113-80090253566-30971527105 | 05161542529-00085727016-35134775864-52531204064-98744439797 |
    |  7 | 5014 | 33071042495-29920376648-91343430102-79082003121-73317691963-02846712788-88069761578-14885283975-44409837760-90760298045 | 91798303270-64988107984-08161247972-12116454627-22996445111 |
    |  8 | 5000 | 73754818686-04889373966-18668178968-56957589012-31352882173-91882653509-59577900152-88962682169-52981807259-62646890059 | 76460662325-41613089656-42706083314-81833284991-17063140920 |
    |  9 | 5273 | 26482547570-00155460224-12388481921-23289186371-78242522654-77998886134-73270876420-50821093220-31442690639-11588920653 | 30508501104-50823269125-88107014550-70202920684-95842308929 |
    | 10 | 5035 | 05677017559-47107518969-97509137401-28934334557-14497052050-61906823704-44077628507-24840441785-05187301456-27797851637 | 29489382504-13697582598-09964978366-26554639515-36136545002 |
    +----+------+-------------------------------------------------------------------------------------------------------------------------+-------------------------------------------------------------+
    10 rows in set (0.00 sec)
    
    mysql>

# 参考: #

1.  《高性能MySQL》
2.  《千金良方 MySQL性能优化金字塔法则》
3.  http://www.360doc.com/content/12/0809/11/2459\_229177650.shtml\#
4.  https://www.cnblogs.com/lfri/p/10724943.html
5.  https://www.cnblogs.com/bulh/articles/11229473.html
6.  https://blog.51cto.com/u\_14036860/2515630
7.  http://blog.itpub.net/28536251/viewspace-2139708/
8.  https://www.cnblogs.com/shouke/p/14618738.html

[http_freecode.com_projects_fio]: http://freecode.com/projects/fio