innodb大表数据测试

有这样一个业务场景，数据的写入就是回流，平常都是依赖user_id查询，数据大概5亿左右，因此我做了下测试，测试过程如下：

性能测试机的环境：

      Uptime | 364 days, 4:02, 2 users, load average: 0.43, 0.19, 0.06
    Platform | Linux
     Release | Red Hat Enterprise Linux Server release 5.4 (Tikanga)
      Kernel | 2.6.18-164.el5
Architecture | CPU = 64-bit, OS = 64-bit
   Threading | NPTL 2.5
    Compiler | GNU CC version 4.1.2 20080704 (Red Hat 4.1.2-44).
     SELinux | Disabled
 Virtualized | No virtualization detected
# Processor ##################################################
  Processors | physical = 2, cores = 8, virtual = 16, hyperthreading = yes
      Speeds | 16x2261.063
      Models | 16xIntel(R) Xeon(R) CPU E5520 @ 2.27GHz
      Caches | 16x8192 KB
# Memory #####################################################
       Total | 23.53G
        Free | 106.74M
        Used | physical = 23.43G, swap = 5.02M, virtual = 23.43G
     Buffers | 165.41M
      Caches | 4.78G
       Dirty | 572 kB
     UsedRSS | 18.3G
  Swappiness | vm.swappiness = 60
 DirtyPolicy | vm.dirty_ratio = 40, vm.dirty_background_ratio = 10

MySQL的版本5.1.48

2.5亿数据测试：

数据准备：

./sysbench --test=oltp --mysql-table-engine=innodb --oltp-table-size=250000000 --mysql-user=lingluo --mysql-password=lingluo --mysql-socket=/u01/mysql/run/mysql.sock prepare

执行：

$./sysbench --test=oltp --mysql-table-engine=innodb --oltp-table-size=250000000 --oltp-read-only --mysql-user=root --mysql-socket=/u01/mysql/run/mysql.sock run
sysbench 0.4.12:  multi-threaded system evaluation benchmark

No DB drivers specified, using mysql
Running the test with following options:
Number of threads: 1

Doing OLTP test.
Running mixed OLTP test
Doing read-only test
Using Special distribution (12 iterations,  1 pct of values are returned in 75 pct cases)
Using "BEGIN" for starting transactions
Using auto_inc on the id column
Maximum number of requests for OLTP test is limited to 10000
Threads started!
Done.

OLTP test statistics:
    queries performed:
        read:                            140000
        write:                           0
        other:                           20000
        total:                           160000
    transactions:                        10000  (61.46 per sec.)
    deadlocks:                           0      (0.00 per sec.)
    read/write requests:                 140000 (860.42 per sec.)
    other operations:                    20000  (122.92 per sec.)

Test execution summary:
    total time:                          162.7105s
    total number of events:              10000
    total time taken by event execution: 162.6381
    per-request statistics:
         min:                                  2.86ms
         avg:                                 16.26ms
         max:                                280.82ms
         approx.  95 percentile:              33.86ms

Threads fairness:
    events (avg/stddev):           10000.0000/0.00
    execution time (avg/stddev):   162.6381/0.00

数据清理：

$./sysbench --test=oltp --mysql-table-engine=innodb --oltp-table-size=250000000 --mysql-user=root --mysql-socket=/u01/mysql/run/mysql.sock cleanup
sysbench 0.4.12:  multi-threaded system evaluation benchmark

No DB drivers specified, using mysql
Dropping table 'sbtest'...
Done.

5亿数据测试：

数据准备：

./sysbench --test=oltp --mysql-table-engine=innodb --oltp-table-size=500000000 --mysql-user=lingluo --mysql-password=lingluo --mysql-socket=/u01/mysql/run/mysql.sock prepare

执行：

$./sysbench --test=oltp --mysql-table-engine=innodb --oltp-table-size=500000000 --oltp-read-only --mysql-user=root --mysql-socket=/u01/mysql/run/mysql.sock run
sysbench 0.4.12:  multi-threaded system evaluation benchmark

No DB drivers specified, using mysql
Running the test with following options:
Number of threads: 1

Doing OLTP test.
Running mixed OLTP test
Doing read-only test
Using Special distribution (12 iterations,  1 pct of values are returned in 75 pct cases)
Using "BEGIN" for starting transactions
Using auto_inc on the id column
Maximum number of requests for OLTP test is limited to 10000
Threads started!
Done.

OLTP test statistics:
    queries performed:
        read:                            140000
        write:                           0
        other:                           20000
        total:                           160000
    transactions:                        10000  (14.27 per sec.)
    deadlocks:                           0      (0.00 per sec.)
    read/write requests:                 140000 (199.83 per sec.)
    other operations:                    20000  (28.55 per sec.)

Test execution summary:
    total time:                          700.6116s
    total number of events:              10000
    total time taken by event execution: 700.5176
    per-request statistics:
         min:                                 12.31ms
         avg:                                 70.05ms
         max:                                445.24ms
         approx.  95 percentile:             110.52ms

Threads fairness:
    events (avg/stddev):           10000.0000/0.00
    execution time (avg/stddev):   700.5176/0.00

清理数据

结论：

5亿的数据，95%的范围还是在111ms之内，稍微慢了点，2亿的数据，95%的查询时间范围在34ms，速度还可以接受。在这个场景下，单表数据量2亿还是可以。