前往小程序,Get更优阅读体验!
立即前往
首页
学习
活动
专区
工具
TVP
发布
社区首页 >专栏 >使用my2sql提取binlog里的数据

使用my2sql提取binlog里的数据

原创
作者头像
保持热爱奔赴山海
修改2023-09-06 16:27:27
1790
修改2023-09-06 16:27:27
举报
文章被收录于专栏:饮水机管理员饮水机管理员

项目地址: https://github.com/liuhr/my2sql

参数和说明

代码语言:javascript
复制
# ./my2sql --help
my2sql V2.0
-U	prefer to use unique key instead of primary key to build where condition for delete/update sql
-add-extraInfo
		Works with -work-type=2sql|rollback. Print database/table/datetime/binlogposition...info on the line before sql, default false
-big-trx-row-limit int
		transaction with affected rows greater or equal to this value is considerated as big transaction. Valid values range from 1 to 30000, default 10 (default 10)
-databases string
		only parse these databases, comma seperated, default all.
-do-not-add-prifixDb
		Prefix table name witch database name in sql,ex: insert into db1.tb1 (x1, x1) values (y1, y1). 
-file-per-table
		One file for one table if true, else one file for all tables. default false. Attention, always one file for one binlog
-full-columns
		For update sql, include unchanged columns. for update and delete, use all columns to build where condition.	
		default false, this is, use changed columns to build set part, use primary/unique key to build where condition
-host string
		mysql host, default 127.0.0.1 . (default "127.0.0.1")
-ignore-databases string
		ignore parse these databases, comma seperated, default null
-ignore-primaryKey-forInsert
		for insert statement when -workType=2sql, ignore primary key
-ignore-tables string
		ignore parse these tables, comma seperated, default null
-local-binlog-file string
		local binlog files to process, It works with -mode=file 
-long-trx-seconds int
		transaction with duration greater or equal to this value is considerated as long transaction. Valid values range from 0 to 3600, default 1 (default 1)
-mode string
		valid options are:  repl,file. repl: as a slave to get binlogs from master. file: get binlogs from local filesystem. default repl (default "repl")
-mysql-type string
		valid options are:  mysql,mariadb. server of binlog, mysql or mariadb, default mysql (default "mysql")
-output-dir string
		result output dir, default current work dir. Attension, result files could be large, set it to a dir with large free space
-output-toScreen
		Just output to screen,do not write to file
-password string
		mysql user password.
-port uint
		mysql port, default 3306. (default 3306)
-print-interval int
		works with -w='stats', print stats info each PrintInterval. Valid values range from 1 to 600, default 30 (default 30)
-server-id uint
		this program replicates from mysql as slave to read binlogs. Must set this server id unique from other slaves, default 1113306 (default 1113306)
-sql string
		valid options are:  insert,update,delete. only parse these types of sql, comma seperated, valid types are: insert, update, delete; default is all(insert,update,delete)
-start-datetime string
		Start reading the binlog at first event having a datetime equal or posterior to the argument, it should be like this: "2020-01-01 01:00:00"
-start-file string
		binlog file to start reading
-start-pos uint
		start reading the binlog at position (default 4)
-stop-datetime string
		Stop reading the binlog at first event having a datetime equal or posterior to the argument, it should be like this: "2020-12-30 01:00:00"
-stop-file string
		binlog file to stop reading
-stop-pos uint
		Stop reading the binlog at position (default 4)
-tables string
		only parse these tables, comma seperated, DONOT prefix with schema, default all.
-threads uint
		Works with -workType=2sql|rollback. threads to run (default 2)
-tl string
		time location to parse timestamp/datetime column in binlog, such as Asia/Shanghai. default Local (default "Local")
-user string
		mysql user. 
-v	print version
-work-type string
		valid options are:  2sql,rollback,stats. 2sql: convert binlog to sqls, rollback: generate rollback sqls, stats: analyze transactions. default: 2sql (default "2sql")


说明: 
	-sql string 提取的变更类型,默认是all,也可以 insert,update,delete 用逗号分隔开
	-add-extraInfo  是否把database/table/datetime/binlogposition...信息以注释的方式加入生成的每条sql前,默认false

	-full-columns   For update sql, include unchanged columns. for update and delete, use all columns to build where condition. default false, this is, use changed columns to build set part, use primary/unique key to build where condition 生成的sql是否带全列信息,默认false  【测试发现,如果加了这个参数,但是解析的表里面有json列,最终生成的rollback.sql里面的where条件的json字段内容会出现解析不正确的情况】		

已知的问题

代码语言:javascript
复制
1、对于bit类型,my2sql生成的是不带b标识的,例如 b'0' 会被转成0 b'1' 会被转成 1 。这个是有点问题的,需要人工修复。
2、对于json列,如果解析的时候加了-full-columns参数,生成的sql的where条件里面的json列的内容拼装会有问题,因此一般不加这个参数。

使用示例

1、分析本地离线binlog文件模式

代码语言:javascript
复制
 1 账号 需要有查看表结构的权限
 2 密码策略必须是 mysql_native_password 模式

 # 提取当时的操作记录
 mkdir -pv tmpdir_offline_forward_0316
 ./my2sql -user dts -password 'dts' -host 192.168.1.11 -port 3306 -mode file -work-type 2sql -start-file ./mysql-bin.001916 -local-binlog-file ./mysql-bin.001916 -databases db1 -tables tb1 -output-dir ./tmpdir_offline_forward_0316

 -rw-r--r-- 1 root root 284 2023-03-08 17:27 biglong_trx.txt
 -rw-r--r-- 1 root root 292 2023-03-08 17:27 binlog_status.txt 
 -rw-r--r-- 1 root root 28K 2023-03-08 17:27 forward.1916.sql  #  提取到的sql明细


 # 反向生成回滚记录
 mkdir -pv tmpdir_offline_rollback
 ./my2sql -user dts -password 'dts' -host 192.168.1.11 -port 3306 -mode file -work-type rollback -start-file ./mysql-bin.001916 -local-binlog-file ./mysql-bin.001916 -databases db1 -tables tb1 -output-dir ./tmpdir_offline_rollback

 -rw-r--r-- 1 root root 284 2023-03-08 17:27 biglong_trx.txt
 -rw-r--r-- 1 root root 292 2023-03-08 17:27 binlog_status.txt
 -rw-r--r-- 1 root root 28K 2023-03-08 17:27 rollback.1916.sql  # 提取到的回滚sql明细

2、伪装成从库,直接分析远程mysql指定的binlog文件模式

代码语言:javascript
复制
 1 需要连接数据库的用户有SELECT, REPLICATION SLAVE, REPLICATION CLIENT权限
 2 使用rollback功能时,要解析的binlog段,表结构要保持一致
 3 密码策略必须是 mysql_native_password 模式

 # 提取当时的操作记录
 mkdir -pv tmpdir_online_forward_0316
 ./my2sql -user dts -password 'dts' -host 192.168.1.11 -port 3306 -mode repl -work-type 2sql -start-file binlog.000014 -start-datetime "2023-03-16 20:00:00" -stop-datetime "2023-03-16 22:00:00" -databases db1 -tables tb1 -output-dir ./tmpdir_online_forward_0316

 # 反向生成回滚记录
 mkdir -pv tmpdir_online_rollback_0316
 ./my2sql -user dts -password 'dts' -host 192.168.1.11 -port 3306 -mode repl -work-type rollback -start-file binlog.000014 -start-datetime "2023-03-16 20:00:00" -stop-datetime "2023-03-16 22:00:00" -databases db1 -tables tb1 -add-extraInfo -output-dir ./tmpdir_online_rollback_0316

原创声明:本文系作者授权腾讯云开发者社区发表,未经许可,不得转载。

如有侵权,请联系 cloudcommunity@tencent.com 删除。

原创声明:本文系作者授权腾讯云开发者社区发表,未经许可,不得转载。

如有侵权,请联系 cloudcommunity@tencent.com 删除。

评论
登录后参与评论
0 条评论
热度
最新
推荐阅读
相关产品与服务
云数据库 MySQL
腾讯云数据库 MySQL(TencentDB for MySQL)为用户提供安全可靠,性能卓越、易于维护的企业级云数据库服务。其具备6大企业级特性,包括企业级定制内核、企业级高可用、企业级高可靠、企业级安全、企业级扩展以及企业级智能运维。通过使用腾讯云数据库 MySQL,可实现分钟级别的数据库部署、弹性扩展以及全自动化的运维管理,不仅经济实惠,而且稳定可靠,易于运维。
领券
问题归档专栏文章快讯文章归档关键词归档开发者手册归档开发者手册 Section 归档