MySQL Casual Talks Vol.3 Lightning Talks Akinori Yoshida @marqs

mitsubachiworks inc.

We love Maatkit

lots of Hatena Bookmarks

Ops guys at Pinterest use Maatkit http://speakerdeck.com/u/yashh/p/scaling-pinterest

Maatkit has become part of Percona Toolkit, and there will be no further development or releases of Maatkit separately from Percona Toolkit. http://www.maatkit.org/

Maatkit Percona Toolkit

Percona Toolkit is derived from Maatkit and Aspersa, two of the best- known toolkits for MySQL server administration. It is developed and supported by Percona Inc. http://www.percona.com/software/percona-toolkit/

some new tools.

pt-mext

Look at many samples of MySQL SHOW GLOBAL STATUS side-by-side. http://www.percona.com/doc/percona-toolkit/2.1/pt-mext.html

server01:~$ pt-mext -r -- mysqladmin ext -i3 -c7 -uxxx -pxxx|grep buffer Innodb_buffer_pool_pages_data 380 0 0 0 Innodb_buffer_pool_pages_dirty 9 2 2 -10 Innodb_buffer_pool_pages_flushed 2958 0 0 0 Innodb_buffer_pool_pages_free 131 0 0 0 Innodb_buffer_pool_pages_misc 1 0 0 0 Innodb_buffer_pool_pages_total 512 0 0 0 Innodb_buffer_pool_read_ahead_rnd 1 0 0 0 Innodb_buffer_pool_read_ahead_seq 0 0 0 0 Innodb_buffer_pool_read_requests 458289610 1801300 1720438 1144393 Innodb_buffer_pool_reads 12 0 0 0 Innodb_buffer_pool_wait_free 0 0 0 0 Innodb_buffer_pool_write_requests 386698 831 753 520

pt-online-schema-change

@kazeburo

pt-stalk

How do you solve intermittent MySQL problems when you can’t observe them? That’s why pt-stalk exists. http://www.percona.com/doc/percona-toolkit/2.1/pt-stalk.html

define the trigger with the --function, --variable, and -- threshold options http://www.percona.com/doc/percona-toolkit/2.1/pt-stalk.html

server01:~$ pt-stalk --cycles 2 --function processlist --variable State --match update --threshold 10 --dest /var/tmp/ -- -uxxx -pxxx 2012_04_18_16_24_49 Starting /usr/bin/pt-stalk --function=processlist -- variable=State --threshold=10 --match=update --cycles=2 --interval=1 -- iterations= --run-time=30 --sleep=300 --dest=/var/tmp/ --prefix= --notify-by- email= --log=/var/log/pt-stalk.log --pid=/var/run/pt-stalk.pid 2012_04_18_16_24_50 Check results: State=6, matched=no, cycles_true=0 2012_04_18_16_24_55 Check results: State=78, matched=yes, cycles_true=1 2012_04_18_16_24_57 Check results: State=43, matched=yes, cycles_true=2 2012_04_18_16_24_58 Collect triggered 2012_04_18_16_25_01 Collector PID 14808 2012_04_18_16_25_01 Sleeping 300 seconds after collect pt-stalk

server01:/var/tmp$ ll total 2140 drwxrwxrwt 2 root root 4096 2012-04-18 16:27 ./ drwxr-xr-x 12 root root 4096 2012-02-22 05:54 ../ -rw-r--r-- 1 root root 10500 2012-04-18 16:26 2012_04_18_16_24_58-df -rw-r--r-- 1 root root 133 2012-04-18 16:26 2012_04_18_16_24_58-disk-space -rw-r--r-- 1 root root 33210 2012-04-18 16:26 2012_04_18_16_24_58-diskstats -rw-r--r-- 1 root root 17 2012-04-18 16:26 2012_04_18_16_24_58-hostname -rw-r--r-- 1 root root 34616 2012-04-18 16:25 2012_04_18_16_24_58- innodbstatus1 -rw-r--r-- 1 root root 34801 2012-04-18 16:26 2012_04_18_16_24_58- innodbstatus2 -rw-r--r-- 1 root root 31320 2012-04-18 16:27 2012_04_18_16_24_58-interrupts -rw-r--r-- 1 root root 7425 2012-04-18 16:25 2012_04_18_16_24_58-iostat -rw-r--r-- 1 root root 563 2012-04-18 16:25 2012_04_18_16_24_58-iostat- overall -rw-r--r-- 1 root root 4215 2012-04-18 16:25 2012_04_18_16_24_58-log_error -rw-r--r-- 1 root root 12620 2012-04-18 16:25 2012_04_18_16_24_58-lsof -rw-r--r-- 1 root root 36420 2012-04-18 16:26 2012_04_18_16_24_58-meminfo -rw-r--r-- 1 root root 8380 2012-04-18 16:25 2012_04_18_16_24_58-mpstat -rw-r--r-- 1 root root 608 2012-04-18 16:25 2012_04_18_16_24_58-mpstat- overall -rw-r--r-- 1 root root 73055 2012-04-18 16:25 2012_04_18_16_24_58-mutex- status1 -rw-r--r-- 1 root root 73061 2012-04-18 16:26 2012_04_18_16_24_58-mutex- status2 -rw-r--r-- 1 root root 460229 2012-04-18 16:25 2012_04_18_16_24_58-mysqladmin -rw-r--r-- 1 root root 20580 2012-04-18 16:26 2012_04_18_16_24_58-netstat

server01:/var/tmp$ pt-sift /var/tmp/ ======== server01 at 2012_04_18_16_24_58 DEFAULT (1 of 1) ======== --diskstats-- #ts device rd_s rd_avkb rd_mb_s rd_mrg rd_cnc rd_rt wr_s wr_avkb wr_mb_s wr_mrg wr_cnc wr_rt busy in_prg io_s qtime stime {29} xvda1 0.2 8.7 0.0 0% 0.0 4.7 31.7 7.0 0.2 31% 0.3 6.1 24% 0 31.9 0.8 5.2 xvda1 0% 35% 15% 30% 0% 65% 70% 65% . 50% 55% 60% 50% 60% . 50% 65% 55% 45% 65% 40% 60% 80% 5% 0% 15% 0% 25% 10% 0% --vmstat-- r b swpd free buff cache si so bi bo in cs us sy id wa 51 4 0 325532 97472 52648 0 0 6 36 15 328 2 1 84 2 16 1 0 320592 97552 53708 0 0 17 44 18 161 1 2 0 4 wa 0% . . . . . . . . . . . . . . 40% 0% . . . 10% 60% 50% 45% . . 50% 30% 45% 30% --innodb-- txns: 92xACTIVE (50s) 8xCOMMITTED (0s) 2xnot (0s) 8 queries inside InnoDB, 92 queries in queue Main thread: flushing log, pending reads 0, writes 1, flush 1 Log: lsn = 0, chkp = 0, chkp age = 0 Threads are waiting at: Threads are waiting on: --processlist-- State 91 Sending data 7 freeing items 3 1 update

very powerful if we can set appropriate trigger :)

one big drawback

Percona Toolkit does *not* have a tool correspond to mk-slave-move

http://www.percona.com/software/percona-toolkit/

thanks!!