我們需要知道的是sysbench並不是一個壓力測試工具，是一個基准測試工具。linux自帶的版本比較低，我們需要自己安裝sysbench。
[root@test2 ~]# sysbench --version sysbench 0.4.12
安裝sysbench，sysbench的源碼托管在GitHub上，下載源碼：
unzip sysbench-master.zip #解壓源碼 yum -y install make automake libtool pkgconfig libaio-devel #下載依賴包 cd sysbench-master sh autogen.sh 編譯： ./configure --with-mysql-includes=/usr/local/mysql/include --with-mysql-libs=/usr/local/mysql/lib #根據安裝的MySQL的位置，設置目錄位置 make make install 這樣安裝之后使用sysbench命令時會報錯。 [root@test3 sysbench-master]# sysbench --version sysbench: error while loading shared libraries: libmysqlclient.so.20: cannot open shared object file: No such file or directory 解決辦法： 在/etc/profile文件中加入一行： export LD_LIBRARY_PATH=/usr/local/mysql/lib source /etc/profile 命令可以正常使用 [root@test3 sysbench-master]# sysbench --version sysbench 1.1.0
查看sysbench的一些幫助信息：
[root@test3 ~]# 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] #線程的數量，默認是1 --events=N limit for total number of events [0] #限制的最大事件數量，默認是0，不限制 --time=N limit for total execution time in seconds [10] #整個測試執行的時間 --warmup-time=N #在進行基准測試多少秒之后啟用統計信息--forced-shutdown=STRING #超過--time時間限制后，強制中斷，默認是【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 #打印出中間的信念，N表示每隔N秒打印一次，0表示禁用--report-checkpoints=[LIST,...] #轉儲完全統計信息並在指定時間點復位所有計數器，參數是逗號分隔值的列表，表示從必須執行報告檢查點的測試開始所經過的時間（以秒為單位）。 默認情況下，報告檢查點處於關閉狀態[off]。--debug[=on|off] print more debugging info [off] --validate[=on|off] #在可能情況下執行驗證檢查，默認是[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
#上面是一些通用的配置信息，在具體測試某個測試時，會再詳細說明參數設置
首先來進行IO測試
[root@test3 ~]# sysbench fileio help #查看IO測試的文檔 sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3) fileio options: --file-num=N number of files to create [128] #文件的數量 --file-block-size=N block size to use in all IO operations [16384] #文件塊的大小，如果要是針對INNODB的測試，可以設置為innodb_page_size的大小 --file-total-size=SIZE total size of files to create [2G] #文件的總大小 --file-test-mode=STRING test mode {seqwr【順序寫】, seqrewr【順序讀寫】, seqrd【順序讀】, rndrd【隨機讀】, rndwr【隨機寫】, rndrw【隨機讀寫】} #文件測試模式 --file-io-mode=STRING file operations mode {sync【同步】,async【異步】,mmap【map映射】} [默認為：sync] #文件的io模式 --file-async-backlog=N number of asynchronous operatons to queue per thread [128] #打開文件時的選項，這是與API相關的參數。 --file-extra-flags=[LIST,...] #打開文件時的選項，這是與API相關的參數。可選有sync，dsync，direct。--file-fsync-freq=N #執行fsync函數的頻率，fsync主要是同步磁盤文件，因為可能有系統和磁盤緩沖的關系。默認為100，如果為0表示不使用fsync。 --file-fsync-all[=on|off] #每執行完一次寫操作，就執行一次fsync，默認未off。--file-fsync-end[=on|off] #在測試結束時，執行fsync，默認為on。--file-fsync-mode=STRING #文件同步函數的選擇，同樣是和API相關的參數，由於多個操作對fdatasync支持的不同，因此不建議使用fdatasync。默認為fsync。--file-merged-requests=N #盡可能合並此數量的io請求(0-不合並)，默認為[0]。 --file-rw-ratio=N #測試時的讀寫比例，默認是2:1。
在使用sysbench進行測試的時候，通常分為三個步驟prepare,run,cleanup階段。
第一步准備數據（prepare階段）：
[root@test3 systext]# sysbench fileio --file-num=10 --file-total-size=50G prepare sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3) 10 files, 5242880Kb each, 51200Mb 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 53687091200 bytes written in 489.55 seconds (104.59 MiB/sec).
#這里給出一個每秒寫入的數據量104.59MB/s, 這里的寫入是順序寫入的，表示磁盤的吞吐量為104.59MB/s。
【一般對順序的讀寫稱為吞吐量，對隨機的IO使用IOPS來表示】
[root@test3 systext]# ll -h #文件大小為5個G
total 50G
-rw------- 1 root root 5.0G Nov 27 09:30 test_file.0
-rw------- 1 root root 5.0G Nov 27 09:31 test_file.1
-rw------- 1 root root 5.0G Nov 27 09:32 test_file.2
-rw------- 1 root root 5.0G Nov 27 09:32 test_file.3
-rw------- 1 root root 5.0G Nov 27 09:33 test_file.4
-rw------- 1 root root 5.0G Nov 27 09:34 test_file.5
-rw------- 1 root root 5.0G Nov 27 09:35 test_file.6
-rw------- 1 root root 5.0G Nov 27 09:36 test_file.7
-rw------- 1 root root 5.0G Nov 27 09:36 test_file.8
-rw------- 1 root root 5.0G Nov 27 09:37 test_file.9
數據准備好之后，進行測試：
#這里進行隨機讀寫測試
[root@test3 systext]# sysbench fileio --file-num=10 --file-total-size=50G --file-block-size=16384 --file-test-mode=rndrw --file-io-mode=sync --file-extra-flags=direct --time=100 --threads=16 --report-interval=10 run sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3) Running the test with following options: #設定的一些參數數值 Number of threads: 16 Report intermediate results every 10 second(s) Initializing random number generator from current time Extra file open flags: directio 10 files, 5GiB each 50GiB total file size Block size 16KiB Number of IO requests: 0 Read/Write ratio for combined random IO test: 1.50 Periodic FSYNC enabled, calling fsync() each 100 requests. Calling fsync() at the end of test, Enabled. Using synchronous I/O mode Doing random r/w test Initializing worker threads... Threads started! [ 10s ] reads: 3.24 MiB/s writes: 2.16 MiB/s fsyncs: 34.08/s latency (ms,95%): 80.025 #每隔10s輸出一次報告 [ 20s ] reads: 3.49 MiB/s writes: 2.32 MiB/s fsyncs: 36.70/s latency (ms,95%): 73.135 [ 30s ] reads: 3.45 MiB/s writes: 2.29 MiB/s fsyncs: 37.00/s latency (ms,95%): 75.817 [ 40s ] reads: 3.43 MiB/s writes: 2.29 MiB/s fsyncs: 36.00/s latency (ms,95%): 75.817 [ 50s ] reads: 3.57 MiB/s writes: 2.38 MiB/s fsyncs: 37.40/s latency (ms,95%): 73.135 [ 60s ] reads: 3.08 MiB/s writes: 2.06 MiB/s fsyncs: 32.30/s latency (ms,95%): 86.002 [ 70s ] reads: 3.41 MiB/s writes: 2.27 MiB/s fsyncs: 36.40/s latency (ms,95%): 75.817 [ 80s ] reads: 3.47 MiB/s writes: 2.31 MiB/s fsyncs: 36.20/s latency (ms,95%): 73.135 [ 90s ] reads: 3.46 MiB/s writes: 2.31 MiB/s fsyncs: 36.20/s latency (ms,95%): 77.194 [ 100s ] reads: 3.10 MiB/s writes: 2.07 MiB/s fsyncs: 33.50/s latency (ms,95%): 75.817 Throughput: read: IOPS=215.57 3.37 MiB/s (3.53 MB/s) #通常的機械磁盤隨機IOPS也就是200多一點。 write: IOPS=143.72 2.25 MiB/s (2.35 MB/s) #隨機寫入的速度明顯要低很多。 fsync: IOPS=37.13 Latency (ms): min: 0.08 avg: 40.51 max: 1000.31 95th percentile: 77.19 sum: 1601329.71
#隨機讀大概是2.10M/s,文件塊的大小為16KB,可以大概估計磁盤轉速： 2.10*1024KB*60s/16KB=7560n/m, 大概就是7500轉每分
[root@test3 systext]# sysbench fileio --file-num=10 --file-total-size=50G --file-block-size=16384 --file-test-mode=seqrd --file-io-mode=sync --file-extra-flags=direct --time=100 --threads=16 --report-interval=10 run sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3) Running the test with following options: Number of threads: 16 Report intermediate results every 10 second(s) Initializing random number generator from current time Extra file open flags: directio 10 files, 5GiB each 50GiB total file size Block size 16KiB Periodic FSYNC enabled, calling fsync() each 100 requests. Calling fsync() at the end of test, Enabled. Using synchronous I/O mode Doing sequential read test Initializing worker threads... Threads started! [ 10s ] reads: 98.88 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 3.020 [ 20s ] reads: 98.64 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 3.681 [ 30s ] reads: 93.24 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 2.913 [ 40s ] reads: 89.12 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 4.028 [ 50s ] reads: 93.17 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 4.487 [ 60s ] reads: 91.98 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 4.652 [ 70s ] reads: 97.08 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 3.425 [ 80s ] reads: 93.71 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 3.020 [ 90s ] reads: 94.63 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 3.304 [ 100s ] reads: 89.57 MiB/s writes: 0.00 MiB/s fsyncs: 0.00/s latency (ms,95%): 3.364 Throughput: read: IOPS=6016.01 94.00 MiB/s (98.57 MB/s) write: IOPS=0.00 0.00 MiB/s (0.00 MB/s) fsync: IOPS=0.00 Latency (ms): min: 0.40 avg: 2.66 max: 687.00 95th percentile: 3.62 sum: 1599247.42 #測試結果可以看到順序的讀和隨機讀的差距還是超大的
可以更改--file-test-mode的模式，改變測試的模式。
測試階段完成之后，需要進行最后的cleanup階段，
[root@test3 systext]# sysbench fileio --file-num=10 --file-total-size=50 cleanup sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3) Removing test files... [root@test3 systext]# ls [root@test3 systext]# df -h Filesystem Size Used Avail Use% Mounted on /dev/vda3 29G 8.4G 20G 31% / tmpfs 3.9G 44K 3.9G 1% /dev/shm /dev/vda1 190M 30M 151M 17% /boot /dev/vdb 100G 25G 76G 25% /data cgroup_root 3.9G 0 3.9G 0% /cgroup #看到磁盤空間已經釋放
測試MySQL的OLTP
sysbench新版的用法和之前的舊版本有所不同，先來看測試數據庫時的一些參數：
General database options: --db-driver=STRING specifies database driver to use ('help' to get list of available drivers) [mysql] #指定數據庫驅動，默認是mysql --db-ps-mode=STRING prepared statements usage mode {auto, disable} [auto] # --db-debug[=on|off] print database-specific debug information [off] #dubug模式 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] #數據庫名字，默認是sbtest --mysql-ssl[=on|off] use SSL connections, if available in the client library [off] #以下是ssl的連接測試 --mysql-ssl-key=STRING path name of the client private key file --mysql-ssl-ca=STRING path name of the CA file --mysql-ssl-cert=STRING path name of the client public key certificate file --mysql-ssl-cipher=STRING use specific cipher for SSL connections [] --mysql-compression[=on|off] use compression, if available in the client library [off] #壓縮測試 --mysql-debug[=on|off] trace all client library calls [off] --mysql-ignore-errors=[LIST,...] list of errors to ignore, or "all" [1213,1020,1205] #忽略的錯誤 --mysql-dry-run[=on|off] Dry run, pretend that all MySQL client API calls are successful without executing them [off]
MySQL測試的lua腳本：
#因為是源碼安裝，索引目錄在這里 [root@test3 lua]# pwd /data/sysbench-master/src/lua [root@test3 lua]# ls bulk_insert.lua Makefile oltp_common.lua oltp_point_select.lua oltp_update_index.lua prime-test.lua empty-test.lua Makefile.am oltp_delete.lua oltp_read_only.lua oltp_update_non_index.lua select_random_points.lua internal Makefile.in oltp_insert.lua oltp_read_write.lua oltp_write_only.lua select_random_ranges.lua #根據腳本的名字可以選擇對應的基本
#查看某個lua腳本的用法
[root@test3 lua]# sysbench oltp_common.lua help
sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3)
oltp_common.lua options:
--auto_inc[=on|off] Use AUTO_INCREMENT column as Primary Key (for MySQL), or its alternatives in other DBMS. When disabled, use client-generated IDs [on]
--create_secondary[=on|off] Create a secondary index in addition to the PRIMARY KEY [on]
--create_table_options=STRING Extra CREATE TABLE options []
--delete_inserts=N Number of DELETE/INSERT combinations per transaction [1]
--distinct_ranges=N Number of SELECT DISTINCT queries per transaction [1]
--index_updates=N Number of UPDATE index queries per transaction [1]
--mysql_storage_engine=STRING Storage engine, if MySQL is used [innodb]
--non_index_updates=N Number of UPDATE non-index queries per transaction [1]
--order_ranges=N Number of SELECT ORDER BY queries per transaction [1]
--pgsql_variant=STRING Use this PostgreSQL variant when running with the PostgreSQL driver. The only currently supported variant is 'redshift'. When enabled, create_secondary is automatically disabled, and delete_inserts is set to 0
--point_selects=N Number of point SELECT queries per transaction [10]
--range_selects[=on|off] Enable/disable all range SELECT queries [on]
--range_size=N Range size for range SELECT queries [100]
--secondary[=on|off] Use a secondary index in place of the PRIMARY KEY [off]
--simple_ranges=N Number of simple range SELECT queries per transaction [1]
--skip_trx[=on|off] Don't start explicit transactions and execute all queries in the AUTOCOMMIT mode [off]
--sum_ranges=N Number of SELECT SUM() queries per transaction [1]
--table_size=N Number of rows per table [10000]
--tables=N Number of tables [1]
prepare階段：
創建默認的測試庫：
mysql> create database sbtest; #創建數據庫 Query OK, 1 row affected (0.11 sec) #准備數據，時間比較長，可以把table_size設置的小一點 [root@test3 lua]# sysbench /data/sysbench-master/src/lua/oltp_read_write.lua --tables=3 --table_size=10000000 --mysql-user=root --mysql-password=123456 --mysql-host=10.0.102.214 --mysql-port=3306 --mysql-db=sbtest prepare sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3) Creating table 'sbtest1'... Inserting 10000000 records into 'sbtest1' Creating a secondary index on 'sbtest1'... Creating table 'sbtest2'... Inserting 10000000 records into 'sbtest2' Creating a secondary index on 'sbtest2'... Creating table 'sbtest3'... Inserting 10000000 records into 'sbtest3' Creating a secondary index on 'sbtest3'...
#在MySQL shel1中查看數據
mysql> select count(*) from sbtest1;
+----------+
| count(*) |
+----------+
| 10000000 |
+----------+
1 row in set (1.89 sec)
mysql> show tables;
+------------------+
| Tables_in_sbtest |
+------------------+
| sbtest1 |
| sbtest2 |
| sbtest3 |
+------------------+
3 rows in set (0.00 sec)
run階段
選擇一個合適的lua腳本進行測試：
[root@test3 lua]# sysbench /data/sysbench-master/src/lua/oltp_point_select.lua --tables=3 --table_size=10000000 --mysql-user=root --mysql-password=123456 --mysql-host=10.0.102.214 --mysql-port=3306 --mysql-db=sbtest --threads=128 --time=100 --report-interval=5 run sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3) Running the test with following options: Number of threads: 128 Report intermediate results every 5 second(s) Initializing random number generator from current time Initializing worker threads... Threads started! [ 5s ] thds: 128 tps: 15037.47 qps: 15037.47 (r/w/o: 15037.47/0.00/0.00) lat (ms,95%): 41.10 err/s: 0.00 reconn/s: 0.00 [ 10s ] thds: 128 tps: 18767.43 qps: 18767.43 (r/w/o: 18767.43/0.00/0.00) lat (ms,95%): 46.63 err/s: 0.00 reconn/s: 0.00 [ 15s ] thds: 128 tps: 22463.68 qps: 22463.68 (r/w/o: 22463.68/0.00/0.00) lat (ms,95%): 40.37 err/s: 0.00 reconn/s: 0.00 [ 20s ] thds: 128 tps: 26848.42 qps: 26848.42 (r/w/o: 26848.42/0.00/0.00) lat (ms,95%): 28.67 err/s: 0.00 reconn/s: 0.00 [ 25s ] thds: 128 tps: 27005.57 qps: 27005.57 (r/w/o: 27005.57/0.00/0.00) lat (ms,95%): 15.00 err/s: 0.00 reconn/s: 0.00 [ 30s ] thds: 128 tps: 26965.62 qps: 26965.62 (r/w/o: 26965.62/0.00/0.00) lat (ms,95%): 1.82 err/s: 0.00 reconn/s: 0.00 [ 35s ] thds: 128 tps: 27626.74 qps: 27626.74 (r/w/o: 27626.74/0.00/0.00) lat (ms,95%): 0.42 err/s: 0.00 reconn/s: 0.00 [ 40s ] thds: 128 tps: 27244.27 qps: 27244.27 (r/w/o: 27244.27/0.00/0.00) lat (ms,95%): 0.33 err/s: 0.00 reconn/s: 0.00 [ 45s ] thds: 128 tps: 26522.56 qps: 26522.56 (r/w/o: 26522.56/0.00/0.00) lat (ms,95%): 1.42 err/s: 0.00 reconn/s: 0.00 [ 50s ] thds: 128 tps: 26791.43 qps: 26791.43 (r/w/o: 26791.43/0.00/0.00) lat (ms,95%): 5.57 err/s: 0.00 reconn/s: 0.00 [ 55s ] thds: 128 tps: 27088.42 qps: 27088.42 (r/w/o: 27088.42/0.00/0.00) lat (ms,95%): 1.42 err/s: 0.00 reconn/s: 0.00 [ 60s ] thds: 128 tps: 28056.06 qps: 28056.06 (r/w/o: 28056.06/0.00/0.00) lat (ms,95%): 0.22 err/s: 0.00 reconn/s: 0.00 [ 65s ] thds: 128 tps: 27296.11 qps: 27296.11 (r/w/o: 27296.11/0.00/0.00) lat (ms,95%): 0.73 err/s: 0.00 reconn/s: 0.00 [ 70s ] thds: 128 tps: 28621.60 qps: 28621.60 (r/w/o: 28621.60/0.00/0.00) lat (ms,95%): 0.19 err/s: 0.00 reconn/s: 0.00 [ 75s ] thds: 128 tps: 28992.29 qps: 28992.29 (r/w/o: 28992.29/0.00/0.00) lat (ms,95%): 0.19 err/s: 0.00 reconn/s: 0.00 [ 80s ] thds: 128 tps: 28279.88 qps: 28279.88 (r/w/o: 28279.88/0.00/0.00) lat (ms,95%): 0.20 err/s: 0.00 reconn/s: 0.00 [ 85s ] thds: 128 tps: 28612.84 qps: 28612.84 (r/w/o: 28612.84/0.00/0.00) lat (ms,95%): 0.20 err/s: 0.00 reconn/s: 0.00 [ 90s ] thds: 128 tps: 28031.47 qps: 28031.47 (r/w/o: 28031.47/0.00/0.00) lat (ms,95%): 0.20 err/s: 0.00 reconn/s: 0.00 [ 95s ] thds: 128 tps: 28734.66 qps: 28734.66 (r/w/o: 28734.66/0.00/0.00) lat (ms,95%): 0.20 err/s: 0.00 reconn/s: 0.00 [ 100s ] thds: 128 tps: 28767.20 qps: 28767.20 (r/w/o: 28767.20/0.00/0.00) lat (ms,95%): 2.39 err/s: 0.00 reconn/s: 0.00 SQL statistics: queries performed: read: 2638920 #總的select數量 write: 0 other: 0 total: 2638920 transactions: 2638920 (26382.71 per sec.) #TPS queries: 2638920 (26382.71 per sec.) #QPS ignored errors: 0 (0.00 per sec.) #忽略的錯誤 reconnects: 0 (0.00 per sec.) #重新連接 Throughput: events/s (eps): 26382.7081 #每秒的事件數，一般和TPS一樣 time elapsed: 100.0246s #測試的總時間 total number of events: 2638920 #總的事件數，一般和TPS一樣 Latency (ms): min: 0.11 #最小響應時間 avg: 4.85 #平均響應時間 max: 649.29 #最大響應時間 95th percentile: 25.74 #95%的響應時間是這個數據 sum: 12796148.28 Threads fairness: events (avg/stddev): 20616.5625/196.08 execution time (avg/stddev): 99.9699/0.00
#在這個測試中，可以看到TPS與QPS的大小基本一致，說明這個lua腳本中的一個查詢一般就是一個事務！
我們一般關注的指標主要有:
- response time avg：平均響應時間（后面的95%的大小可以通過–percentile=98的方式去更改）。
- transactions：精確的說是這一項后面的TPS，但如果使用了–skip-trx=on，這項事務數為0，需要用total number of events去除以總時間，得到tps（其實還可以分為讀tps和寫tps）。
- queries：用它除以總時間，得到吞吐量QPS。
因為上面的TPS與QPS是一樣的，因此只繪了TPS的圖，如下：
剛開始的時候有一個明顯的上升，這時候是因為在bp中沒有緩存數據，需要從磁盤中讀數據，也就是預熱階段！
清理數據
[root@test3 lua]# sysbench /data/sysbench-master/src/lua/oltp_read_write.lua --tables=3 --table_size=10000000 --mysql-user=root --mysql-password=123456 --mysql-host=10.0.102.214 --mysql-port=3306 --mysql-db=sbtest cleanup sysbench 1.1.0 (using bundled LuaJIT 2.1.0-beta3) Dropping table 'sbtest1'... Dropping table 'sbtest2'... Dropping table 'sbtest3'... [root@test3 lua]#
sysbench除了以上的測試之外，還可以測試：
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