Recently, the company plans to go to oracle and migrate to mysql cluster. Our technology selection of distributed database middleware chose mycat. This is undoubtedly the preparatory work that needs to be done well. With the help of the operation and maintenance students, the cluster has been set up, and our next task is stress testing. This article simply records the installation and use of the pressure measurement tool Sysbench I have chosen.
Introduction to sysbench
Sysbench is a modular, cross-platform, multi-threaded benchmark test tool, mainly used to evaluate and test database load conditions under various system parameters. It mainly includes the following tests: cpu performance, disk io performance, thread scheduling performance, memory allocation and transmission speed and database performance. Since I am a dba, I focus on the scenario of sysbench testing the database. Currently sysbench supports three data sources: mysql, postgreSQL, and oracle.
test environment
System: Centos-8
Sysbench: sysbench-0.4.12.10, download link: https://github.com/akopytov/sysbench
You can also click to download directly: http://downloads.mysql.com/source/sysbench-0.4.12.10.tar.gz
installation steps
Some dependency installation of sysbench
yum -y install make automake libtool pkgconfig libaio-devel vim-common
如果没有安装mysql,则需要安装mysql-devel
yum install mysql-develInstall sysbench
- Enter the sysbench source directory
cd /data/sysbench/sysbench-0.4.12- Execute autogen.sh and use it to generate the configure file
./autogen.sh- Execute configure && make && make install to complete the sysbench installation
./configure --prefix=/wls/sysbench/ --build=x86_64
make && make installNote: If you do not add --build=x86_64, the following error may be prompted:
checking build system type... Invalid configuration `x86_64-unknown-linux-': machine `x86_64-unknown-linux' not recognized
configure: error: /bin/sh config/config.sub x86_64-unknown-linux- failedVerify that the installation is successful
[root@05823ea529b0 sysbench-1.0.15]# /wls/sysbench/bin/sysbench --version
sysbench 0.4.12.10
[root@05823ea529b0 sysbench-1.0.15]# ./sysbench --test=cpu run
sysbench 0.4.12.10 (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: 10000
Initializing worker threads...
Threads started!
CPU speed:
events per second: 1158.09
General statistics:
total time: 10.0005s
total number of events: 11583
Latency (ms):
min: 0.79
avg: 0.86
max: 1.92
95th percentile: 1.01
sum: 9985.69
Threads fairness:
events (avg/stddev): 11583.0000/0.00
execution time (avg/stddev): 9.9857/0.00sysbench view help
Usage:
sysbench [general-options]... --test=<test-name> [test-options]... command
General options:
--num-threads=N number of threads to use [1]
--max-requests=N limit for total number of requests [10000]
--max-time=N limit for total execution time in seconds [0]
--forced-shutdown=STRING amount of time to wait after --max-time before forcing shutdown [off]
--thread-stack-size=SIZE size of stack per thread [32K]
--init-rng=[on|off] initialize random number generator [off]
--seed-rng=N seed for random number generator, ignored when 0 [0]
--tx-rate=N target transaction rate (tps) [0]
--tx-jitter=N target transaction variation, in microseconds [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. []
--test=STRING test to run
--debug=[on|off] print more debugging info [off]
--validate=[on|off] perform validation checks where possible [off]
--help=[on|off] print help and exit
--version=[on|off] print version and exit
Log options:
--verbosity=N verbosity level {5 - debug, 0 - only critical messages} [4]
--percentile=N percentile rank of query response times to count [95]
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
oltp - OLTP test
Commands: prepare run cleanup help version
See 'sysbench --test=<name> help' for a list of options for each test.
The process of sysbench stress testing the database
- The prepare phase is used to prepare, compare to establish the test table, and fill the table with data;
- This stage of the run phase is the SQL to run the stress test;
- The cleanup stage is to clear the data, that is, all tables initialized in the prepare stage should be dropped;
The test types in sysbench can be roughly divided into built-in, lua script custom tests
- 内置: fileio 、cpu 、memory 、threads 、 mutex
- Lua script custom type: sysbench contains some test scripts in the installation directory:
查看帮忙命令:
[root@izwz909ewdz83smewux7a7z bin]# ./sysbench --help
Usage:
sysbench [general-options]... --test=<test-name> [test-options]... command
General options:
--num-threads=N number of threads to use [1]
--max-requests=N limit for total number of requests [10000]
--max-time=N limit for total execution time in seconds [0]
--forced-shutdown=STRING amount of time to wait after --max-time before forcing shutdown [off]
--thread-stack-size=SIZE size of stack per thread [32K]
--init-rng=[on|off] initialize random number generator [off]
--seed-rng=N seed for random number generator, ignored when 0 [0]
--tx-rate=N target transaction rate (tps) [0]
--tx-jitter=N target transaction variation, in microseconds [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. []
--test=STRING test to run
--debug=[on|off] print more debugging info [off]
--validate=[on|off] perform validation checks where possible [off]
--help=[on|off] print help and exit
--version=[on|off] print version and exit
Log options:
--verbosity=N verbosity level {5 - debug, 0 - only critical messages} [4]
--percentile=N percentile rank of query response times to count [95]
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
oltp - OLTP test
Commands: prepare run cleanup help version
See 'sysbench --test=<name> help' for a list of options for each test.
Test mysql through the lua script that comes with sysbench
- Generate table file
[root@izwz909ewdz83smewux7a7z bin]# ./sysbench --test=oltp --oltp-num-tables=100000 --oltp-num-tables=10 --mysql-db=sysbench --mysql-user=root --mysql-password='mydb' --mysql-host=localhost --mysql-port=3306 --db-driver=mysql --init-rng=on prepare
- Multi-threaded test
[root@izwz909ewdz83smewux7a7z bin]# ./sysbench --test=oltp --oltp-table-size=100000 --mysql-db=sysbench --mysql-user=root --mysql-password='mydb' --mysql-host=localhost --mysql-port=3306 --db-driver=mysql --oltp-read-only=off --max-time=60 --num-threads=8 --report-interval=10 --oltp-dist-type=uniform --max-requests=0 --percentile=99 run
sysbench 0.4.12.10: 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
Doing OLTP test.
Running mixed OLTP test
Using Uniform distribution
Using "BEGIN" for starting transactions
Using auto_inc on the id column
Using 1 test tables
Threads started!
[ 10s] Intermediate results: 8 threads, tps: 582.994023, reads/s: 8233.315586, writes/s: 2940.469852 response time: 48.261058ms (99%)
[ 20s] Intermediate results: 8 threads, tps: 589.090480, reads/s: 8248.666697, writes/s: 2945.952392 response time: 35.829896ms (99%)
[ 30s] Intermediate results: 8 threads, tps: 487.603128, reads/s: 6829.243810, writes/s: 2439.015647 response time: 65.415122ms (99%)
[ 40s] Intermediate results: 8 threads, tps: 487.700557, reads/s: 6830.607800, writes/s: 2439.502786 response time: 78.027721ms (99%)
[ 50s] Intermediate results: 8 threads, tps: 478.852588, reads/s: 6703.936228, writes/s: 2394.262939 response time: 71.711459ms (99%)
[ 60s] Intermediate results: 8 threads, tps: 510.752034, reads/s: 7150.528473, writes/s: 2553.760169 response time: 43.136643ms (99%)
Time limit exceeded, exiting...
(last message repeated 7 times)
Done.
OLTP test statistics:
queries performed:
read: 440076 读总数
write: 157170 写总数
other: 62812 他操作总数(SELECT、INSERT、UPDATE、DELETE之外的操作，例如COMMIT等)
total: 660058 总数
transactions: 31378 (522.91 per sec.) 总事务数(每秒事务数)
deadlocks: 56 (0.93 per sec.)
read/write requests: 597246 (9953.02 per sec.) 读写总数(每秒读写次数)
other operations: 62812 (1046.75 per sec.)
General statistics:
total time: 60.0065s 总耗时
total number of events: 31378 共发生多少事务数
total time taken by event execution: 479.8756 所有事务耗时相加(不考虑并行因素)
response time:
min: 3.77ms 最小耗时
avg: 15.29ms 平均耗时
max: 180.78ms 最长耗时
approx. 99 percentile: 58.12ms 超过99%平均耗时
Threads fairness:
events (avg/stddev): 3922.2500/6.02
execution time (avg/stddev): 59.9845/0.00
- Clean up the test table generated above
[root@izwz909ewdz83smewux7a7z bin]# ./sysbench --test=oltp --oltp-num-tables=100000 --oltp-num-tables=10 --mysql-db=sysbench --mysql-user=root --mysql-password='mydb' --mysql-host=localhost --mysql-port=3306 --db-driver=mysql --init-rng=on cleanup
sysbench 0.4.12.10: multi-threaded system evaluation benchmark
Dropping table 'sbtest'...
Dropping table 'sbtest1'...
Dropping table 'sbtest2'...
Dropping table 'sbtest3'...
Dropping table 'sbtest4'...
Dropping table 'sbtest5'...
Dropping table 'sbtest6'...
Dropping table 'sbtest7'...
Dropping table 'sbtest8'...
Dropping table 'sbtest9'...
Done.
[root@izwz909ewdz83smewux7a7z bin]#
Since the above commands are relatively simple, I believe everyone can understand it. If you have any questions, please leave a message and exchange.
After tuning the database, use sysbench to test OLTP to see if TPS will improve.
Note: The sysbench test is only a benchmark test, and does not represent the performance indicators in an actual enterprise environment.
Personal website: http://coderluo.top
Welcome to pay attention to the author's public account: Fall in love with coding , and will regularly share Java technology work, let boring technology swim!
[This article was first published on this site, please indicate the source for reprinting]: http://coderluo.top/2019/10/13/docker/docker-an-zhuang-centos8-bing-shi-yong-ssh-yuan- cheng-lian-jie/