Sysbench-一種系統效能benchmark

Sysbench-一種系統效能benchmark SysBench是一款開源的、跨平台的、模組化的、多線程的效能測試工具，通過高負載地運行在資料庫上，可以執行CPU/記憶體/線程/IO/資料庫等方面的效能測試。用於評估作業系統的績效參數。 1          sysbench簡介Sysbench使得我們無需採用真正的複雜的資料庫benchmark而擷取系統的效能概況。目前主要支援的資料庫有mysql、oracle、postgresql。可評估的系統績效參數有：1.      Cpu效能2.      記憶體配置與傳輸速度測試3.      互斥體測試4.      檔案I/O效能5.      線程調度效能6.      資料庫服務效能（OLTP基準）本來sysbench主要是為mysql編寫的benchmark。但是現在SysBench將進一步擴充，以支援更多的資料庫backends，distributed benchmarks和第三方外掛程式模組。sysbench按照指定的數量開啟線程，每個線程與mysql建立一個串連，每個線程不停地進行事務操作，開啟事務，然後進行一些查詢、更新、插入操作，提交事務，再開始新的事務；所有的sql只訪問一個表-sbtest，是由sysbench的prepare命令建好的。其中的記錄數，也是在prepare時指定好並建立的。測試結束，會有一個統計結果，包括例如每秒事務數、平均回應時間等等； 2          sysbench下載與安裝1.下載：http://sourceforge.net/projects/sysbench/獲得sysbench-0.4.8.tar.gz源碼。sysbench-0.4.8.tar.gz編譯未通過。2.安裝預設支援MySQL，如果需要測試Oracle/PostgreSQL，則在configure時需要加上–with-oracle或者–with-pgsql參數[root@RedHat sysbench-0.4.8]# ./configure --without-mysql[root@RedHat sysbench-0.4.8]# make[root@RedHat sysbench-0.4.8]# make install 如果需要測試mysql，則需要配置參數選項：./configure--with-mysql-includes=/opt/mysql/include/mysql 表示mysql標頭檔目錄 --with-mysql-libs=/opt/mysql/lib/mysql表示mysql模組目錄如股票缺少libtool請至ftp://ftp.gnu.org/gnu/libtool/下載 3          效能測試與結果分析請查看sysbench協助[root@RedHat sysbench-0.4.8]# sysbenchMissing required command argument.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]  --thread-stack-size=SIZE   size of stack per thread [32K]  --init-rng=[on|off]        initialize random number generator [off]  --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 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 See 'sysbench --test=<name> help' for a list of options for each test. [root@RedHat sysbench-0.4.8]#其中通用參數參數名參數描述--num-threads=N測試時使用的線程數，預設為1--max-requests=N請求數最大現在，預設10000-max-time=N執行時間最大限制，預設為0--thread-stack-size=SIZE每個線程的stack大小，預設是32K--init-rng=[on|off]是否初始化隨機數產生器--test=STRING啟動並執行測試--debug=[on|off]是否列印更多的debug資訊--validate=[on|off]是否執行有效性檢查--help=[on|off]列印協助並退出其中--test=STRING中的STRING可選項有：STRING描述fileio檔案IO測試cpuCPU效能測試memory記憶體配置與傳輸速度測試threads線程子系統效能測試mutex互斥體效能測試oltp資料庫OLTP測試 1. 檔案IO測試 [root@RedHat test]# sysbench --test=fileio helpsysbench v0.4.8:  multi-threaded system evaluation benchmark fileio options:  --file-num=N                  number of files to create [128]  --file-block-size=N           block size to use in all IO operations [16384]  --file-total-size=SIZE        total size of files to create [2G]  --file-test-mode=STRING       test mode {seqwr, seqrewr, seqrd, rndrd, rndwr,        rndrw}  --file-io-mode=STRING   file operations mode {sync,async,fastmmap,slowmmap} [sync]  --file-extra-flags=STRING  additional flags to use on opening files {sync,dsync,direct} []  --file-fsync-freq=N      do fsync() after this number of requests (0 - don't use fsync()) [100]  --file-fsync-all=[on|off]     do fsync() after each write operation [off]  --file-fsync-end=[on|off]     do fsync() at the end of test [on]  --file-fsync-mode=STRING which method to use for synchronization {fsync, fdatasync} [fsync]  --file-merged-requests=N merge at most this number of IO requests if poss ible (0- don't merge) [0]  --file-rw-ratio=N             reads/writes ratio for combined test [1.5] [root@RedHat test]#參數選項說明：參數名描述--file-num=N建立的檔案數目，預設為128個--file-block-size=N所有IO操作中的block大小，預設是16384--file-total-size=SIZE建立的檔案的總大小，預設為2G--file-test-mode=STRING測試的訪問模式，可選一下6個參數：seqwr(順序寫)/seqrewr(順序讀寫)/seqrd(順序讀)/ rndwr(隨機寫)/rndrw(隨機讀寫)/rndrd(隨機讀)；--file-io-mode=STRING檔案操作模式,有以下四種選擇：sync,async,fastmmap,slowmmap預設為sync--file-extra-flags=STRING開啟檔案是的標識，有以下選擇：sync,dsync,direct--file-fsync-freq=N在N個請求之後執行fsync()，預設值為100,0表示不執行fsync()--file-fsync-all=[on|off]是否每一次寫操作之後均執行fsync()--file-fsync-end=[on|off]在測試結束時執行fsync()，預設開啟--file-fsync-mode=STRING同步使用什麼方法：fsync, fdatasync，預設為fsync--file-merged-requests=N儘可能的合并N個請求，0表示不合并，預設為0--file-rw-ratio=N 讀寫比例預設為1.5：1測試案例：建立2個進程，建立的檔案總大小為3G，檔案讀寫入模式為隨機讀。可以擷取讀寫速度~[root@RedHat test]# sysbench --test=fileio --num-threads=2 --file-total-size=512M --file-test-mode=rndrw prepare （準備測試檔案）sysbench v0.4.8:  multi-threaded system evaluation benchmark 128 files, 4096Kb each, 512Mb totalCreating files for the test...[root@RedHat test]# sysbench --test=fileio --num-threads=2 --file-total-size=512M --file-test-mode=rndrw run （運行測試）sysbench v0.4.8:  multi-threaded system evaluation benchmark Running the test with following options:Number of threads: 2 Extra file open flags: 0128 files, 4Mb each512Mb total file sizeBlock size 16KbNumber of random requests for random IO: 10000Read/Write ratio for combined random IO test: 1.50Periodic FSYNC enabled, calling fsync() each 100 requests.Calling fsync() at the end of test, Enabled.Using synchronous I/O modeDoing random r/w testThreads started!Done. Operations performed:  6000 Read, 4000 Write, 12800 Other = 22800 TotalRead 93.75Mb  Written 62.5Mb  Total transferred 156.25Mb  (3.2979Mb/sec)  211.07 Requests/sec executed Test execution summary:    total time:                          47.3783s    total number of events:              10000    total time taken by event execution: 36.6912    per-request statistics:         min:                            0.0000s         avg:                            0.0037s         max:                            0.2687s         approx.  95 percentile:         0.0230s Threads fairness:    events (avg/stddev):           5000.0000/180.00    execution time (avg/stddev):   18.3456/0.19 [root@RedHat test]# sysbench --test=fileio --num-threads=2 --file-total-size=512M --file-test-mode=rndrw cleanup （清除測試檔案）sysbench v0.4.8:  multi-threaded system evaluation benchmark Removing test files...[root@RedHat test]#  2. CPU效能測試Sysbench採用尋找最大素數的方式測試CPU效能：[root@RedHat ~]# sysbench --test=cpu helpsysbench v0.4.8:  multi-threaded system evaluation benchmark cpu options:  --cpu-max-prime=N      upper limit for primes generator [10000] [root@RedHat ~]#參數解析參數名描述--cpu-max-prime=N素數的最大限制，預設為10000；測試案例[root@RedHat ~]# sysbench --test=cpu --cpu-max-prime=20000 runsysbench v0.4.8:  multi-threaded system evaluation benchmarkRunning the test with following options:Number of threads: 1Doing CPU performance benchmarkThreads started!Done.Maximum prime number checked in CPU test: 20000Test execution summary:    total time:                          124.1761s    total number of events:              10000    total time taken by event execution: 123.3658    per-request statistics:         min:                            0.0051s         avg:                            0.0123s         max:                            0.1908s         approx.  95 percentile:         0.0171s Threads fairness:    events (avg/stddev):           10000.0000/0.00    execution time (avg/stddev):   123.3658/0.00[root@RedHat ~]# 3.記憶體效能測試 [root@RedHat ~]# sysbench --test=memory helpsysbench v0.4.8:  multi-threaded system evaluation benchmark memory options:  --memory-block-size=SIZE    size of memory block for test [1K]  --memory-total-size=SIZE    total size of data to transfer [100G]  --memory-scope=STRING       memory access scope {global,local} [global]  --memory-hugetlb=[on|off]   allocate memory from HugeTLB pool [off]  --memory-oper=STRING        type of memory operations {read, write, none} [write]  --memory-access-mode=STRING memory access mode {seq,rnd} [seq] [root@RedHat ~]# 參數名描述--memory-block-size=SIZE記憶體塊大小，預設為1K--memory-total-size=SIZE傳輸的資料量，預設為100G--memory-scope=STRING記憶體訪問是局部local還是全域global，預設為global--memory-hugetlb=[on|off]從hugetlb中分配記憶體--memory-oper=STRING記憶體操作類型read, write, none，預設為寫--memory-access-mode=STRING記憶體訪問模式seq,rnd，預設為seq； [root@RedHat test]# sysbench --test=memory --memory-block-size=4k --memory-total-size=64M run  4.線程效能測試 [root@RedHat ~]# sysbench --test=threads helpsysbench v0.4.8:  multi-threaded system evaluation benchmark threads options:  --thread-yields=N      number of yields to do per request [1000]  --thread-locks=N       number of locks per thread [8] [root@RedHat ~]# 參數名描述--thread-yields=N記憶體塊大小，預設為1K--thread-locks=N傳輸的資料量，預設為100G測試案例[root@RedHat test]# sysbench --test=threads --num-threads=64 --thread-yields=100 --thread-locks=2 runsysbench v0.4.8:  multi-threaded system evaluation benchmark Running the test with following options:Number of threads: 64 Doing thread subsystem performance testThread yields per test: 100 Locks used: 2Threads started!Done. Test execution summary:    total time:                          24.8159s    total number of events:              10000    total time taken by event execution: 1575.5529    per-request statistics:         min:                            0.0002s         avg:                            0.1576s         max:                            2.0173s         approx.  95 percentile:         0.5627s Threads fairness:    events (avg/stddev):           156.2500/12.62    execution time (avg/stddev):   24.6180/0.08 [root@RedHat test]# 5互斥體效能測試 [root@RedHat ~]# sysbench --test=mutex helpsysbench v0.4.8:  multi-threaded system evaluation benchmark mutex options:  --mutex-num=N        total size of mutex array [4096]  --mutex-locks=N      number of mutex locks to do per thread [50000]  --mutex-loops=N      number of empty loops to do inside mutex lock [10000] [root@RedHat ~]# 參數名描述--mutex-num=N互斥體資料大小，預設為4096--mutex-locks=N每個線程上互斥鎖的數目，預設為50000；--mutex-loops=N互斥鎖內空迴圈的數目，預設為10000 測試案例[root@RedHat test]# sysbench --test=mutex --num-threads=16 --mutex-num=1024 --mutex-locks=10000 --mutex-loops=5000 runsysbench v0.4.8:  multi-threaded system evaluation benchmark Running the test with following options:Number of threads: 16 Doing mutex performance testThreads started!Done. Test execution summary:    total time:                          0.0482s    total number of events:              16    total time taken by event execution: 0.2271    per-request statistics:         min:                            0.0015s         avg:                            0.0142s         max:                            0.0312s         approx.  95 percentile:         0.0126s Threads fairness:    events (avg/stddev):           1.0000/0.00    execution time (avg/stddev):   0.0142/0.01[root@RedHat test]# 6 OLTP測試不做分析了