sysbench是一个模块化的、跨平台、多线程基准测试工具，主要用于评估测试各种不同系统参数下的数据库负载情况。
sysbench支持以下几种测试模式：
1、CPU运算性能
2、磁盘IO性能
3、调度程序性能
4、内存分配及传输速度
5、POSIX线程性能
6、数据库性能(OLTP基准测试)
目前sysbench主要支持 mysql,drizzle,pgsql,oracle 等几种数据库。

安装
yum install m4  autoconf  automake libtool  mysql-community-devel
git clone https://github.com/akopytov/sysbench.git
cd /sysbench
./autogen.sh
 ./configure --with-mysql-includes=/usr/local/mysql/include --with-mysql-libs=/usr/local/mysql/lib 
make &&make install
------------------------------------------------------------------------------------------------------
报错
wget http://mirror.cogentco.com/pub/mysql/MySQL-5.7/MySQL-shared-5.7.4_m14-1.el6.i686.rpm
rpm -ivh MySQL-shared-5.7.4_m14-1.el6.x86_64.rpm  

whereis sysbench

OLTP测试前准备
初始化测试库环境（总共10个测试表，每个表 100000 条记录，填充随机生成的数据）：

先创建测试数据库oo create databases foo;

sysbench --db-driver=mysql --mysql-user=root --mysql-password=123456 \
  --mysql-db=foo --range_size=100 \
  --table_size=1000000 --tables=10 --threads=8 --events=0 --time=60 \
  --rand-type=uniform /usr/local/src/sysbench/src/lua/oltp_read_only.lua prepare

OLTP测试
 
sysbench --db-driver=mysql --mysql-user=root --mysql-password=123456 \
  --mysql-db=foo --range_size=100 \
  --table_size=1000000 --tables=10 --threads=8 --events=0 --time=60 \
  --rand-type=uniform /usr/local/src/sysbench/src/lua/oltp_read_only.lua run >./sysbench_oltp.log
即：模拟 对10个表并发OLTP测试，每个表1000万行记录，持续压测时间为 1小时。
真实测试场景中，建议持续压测时长不小于30分钟，否则测试数据可能不具参考意义。
测试结果解读：
===================================================================
sysbench 0.5:  multi-threaded system evaluation benchmark

Running the test with following options:
Number of threads: 8
Report intermediate results every 10 second(s)
Random number generator seed is 0 and will be ignored

Threads started!
-- 每10秒钟报告一次测试结果，tps、每秒读、每秒写、99%以上的响应时长统计
[  10s] threads: 8, tps: 1111.51, reads/s: 15568.42, writes/s: 4446.13, response time: 9.95ms (99%)
[  20s] threads: 8, tps: 1121.90, reads/s: 15709.62, writes/s: 4487.80, response time: 9.78ms (99%)
[  30s] threads: 8, tps: 1120.00, reads/s: 15679.10, writes/s: 4480.20, response time: 9.84ms (99%)
[  40s] threads: 8, tps: 1114.20, reads/s: 15599.39, writes/s: 4456.30, response time: 9.90ms (99%)
[  50s] threads: 8, tps: 1114.00, reads/s: 15593.60, writes/s: 4456.70, response time: 9.84ms (99%)
[  60s] threads: 8, tps: 1119.30, reads/s: 15671.60, writes/s: 4476.50, response time: 9.99ms (99%)
OLTP test statistics:
    queries performed:
        read:                            938224    -- 读总数
        write:                           268064    -- 写总数
        other:                           134032    -- 其他操作总数(SELECT、INSERT、UPDATE、DELETE之外的操作，例如COMMIT等)
        total:                           1340320    -- 全部总数
    transactions:                        67016  (1116.83 per sec.)    -- 总事务数(每秒事务数)
    deadlocks:                           0      (0.00 per sec.)    -- 发生死锁总数
    read/write requests:                 1206288 (20103.01 per sec.)    -- 读写总数(每秒读写次数)
    other operations:                    134032 (2233.67 per sec.)    -- 其他操作总数(每秒其他操作次数)

General statistics:    -- 一些统计结果
    total time:                          60.0053s    -- 总耗时
    total number of events:              67016    -- 共发生多少事务数
    total time taken by event execution: 479.8171s    -- 所有事务耗时相加(不考虑并行因素)
    response time:    -- 响应时长统计
         min:                                  4.27ms    -- 最小耗时
         avg:                                  7.16ms    -- 平均耗时
         max:                                 13.80ms    -- 最长耗时
         approx.  99 percentile:               9.88ms    -- 超过99%平均耗时

Threads fairness:
    events (avg/stddev):           8377.0000/44.33
    execution time (avg/stddev):   59.9771/0.00

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]
  --warmup-time=N                 execute events for this many seconds with statistics disabled before the actual benchmark run with statistics enabled [0]
  --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]
  --thread-init-timeout=N         wait time in seconds for worker threads to initialize [30]
  --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
  --luajit-cmd=STRING             perform LuaJIT control command. This option is equivalent to 'luajit -j'. See LuaJIT documentation for more information
  --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, zipfian} to use by default [special]
  --rand-seed=N        seed for random number generator. When 0, the current time is used as an RNG seed. [0]
  --rand-spec-iter=N   number of iterations for the special distribution [12]
  --rand-spec-pct=N    percentage of the entire range where 'special' values will fall in the special distribution [1]
  --rand-spec-res=N    percentage of 'special' values to use for the special distribution [75]
  --rand-pareto-h=N    shape parameter for the Pareto distribution [0.2]
  --rand-zipfian-exp=N shape parameter (exponent, theta) for the Zipfian distribution [0.8]

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)
  --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

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]

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.