整理了Go制作的迁移工具
相对比较
-
- CLIで使える
- MySQLが使える
名前Star数スキーマ方式Dry Run設定liamstask/goose-Go/SQLUp/Down-YAMLmattes/migrate892SQLUp/Down–rubenv/sql-migrate576SQLUp/Down✔YAMLelwinar/rambler371SQLUp/Down-JSONnaoina/migu29Go(struct)Sync✔-dev-cloverlab/carpenter12JSON/実DBSync✔-
2017年1月的时候
本次比较没有考虑的内容。
-
- https://github.com/tanel/dbmigrate
-
- https://github.com/BurntSushi/migration
-
- https://github.com/Boostport/migration
-
- https://github.com/github/gh-ost
-
- https://github.com/wawandco/transporter
-
- https://github.com/GuiaBolso/darwin
-
- https://github.com/DavidHuie/gomigrate
-
- https://github.com/go-gormigrate/gormigrate
-
- https://github.com/pravasan/pravasan
- https://github.com/eure/kamimai
liamstask/goose => liamstask/goose
处理数据库
-
- MySQL
-
- PostgreSQL
- SQLite3
安装
go get bitbucket.org/liamstask/goose/cmd/goose
设定
development:
driver: mysql
open: user=root dbname=goose_test
杂感
-
- 普通な感じ
-
- 使ってる人多そう
- 適用履歴がテーブルに残るのは良い
迁移文件
$ goose create CreatePosts sql
goose: created /Users/nownabe/src/github.com/nownabe/research-go/migration-tools/goose/db/migrations/20170111152547_CreatePosts.sql
使用CLI创建模板并将Up/Down的SQL写入到文件中。
-- +goose Up
-- SQL in section 'Up' is executed when this migration is applied
CREATE TABLE `posts` (
`post_id` INT NOT NULL PRIMARY KEY AUTO_INCREMENT,
`author` VARCHAR(255) NOT NULL,
`text` TEXT NOT NULL
) ENGINE = InnoDB;
-- +goose Down
-- SQL section 'Down' is executed when this migration is rolled back
DROP TABLE `posts`;
虽然Go语言也可以编写迁移文件,但由于似乎不太需要,所以省略。
命令
$ goose
goose is a database migration management system for Go projects.
Usage:
goose [options] <subcommand> [subcommand options]
Options:
-env string
which DB environment to use (default "development")
-path string
folder containing db info (default "db")
-pgschema string
which postgres-schema to migrate (default = none)
Commands:
up Migrate the DB to the most recent version available
down Roll back the version by 1
redo Re-run the latest migration
status dump the migration status for the current DB
create Create the scaffolding for a new migration
dbversion Print the current version of the database
版本管理
保存在名为goose_db_version的表中。
> select * from goose_db_version;
+----+----------------+------------+---------------------+
| id | version_id | is_applied | tstamp |
+----+----------------+------------+---------------------+
| 1 | 0 | 1 | 2017-01-11 15:35:03 |
| 2 | 20170111152547 | 1 | 2017-01-11 15:47:34 |
| 3 | 20170111152547 | 0 | 2017-01-11 15:47:36 |
| 4 | 20170111152547 | 1 | 2017-01-11 15:47:36 |
| 5 | 20170111152547 | 0 | 2017-01-11 15:47:53 |
| 6 | 20170111152547 | 1 | 2017-01-11 15:48:06 |
+----+----------------+------------+---------------------+
6 rows in set (0.00 sec)
[粤语] 迁移/迁徙
处理数据库
-
- Cassandra
-
- MongoDB
-
- MySQL (experimental)
-
- PostgreSQL
-
- SQLite
- Ql
安装
go get -u github.com/mattes/migrate
感想
-
- コマンドが豊富
- CLIの結果が色付きで見やすい
迁移文件
$ migrate -url="mysql://root@tcp(127.0.0.1:3306)/migrate_test" create create_posts
Version 1484119339 migration files created in /Users/nownabe/src/github.com/nownabe/research-go/migration-tools/goose:
1484119339_create_posts.up.sql
1484119339_create_posts.down.sql
在CLI中,将会创建上升/下降两种模板文件。
CREATE TABLE `posts` (
`post_id` INT NOT NULL PRIMARY KEY AUTO_INCREMENT,
`author` VARCHAR(255) NOT NULL,
`text` TEXT NOT NULL
) ENGINE = InnoDB;
DROP TABLE `posts`;
命令
$ migrate
usage: migrate [-path=<path>] -url=<url> <command> [<args>]
Commands:
create <name> Create a new migration
up Apply all -up- migrations
down Apply all -down- migrations
reset Down followed by Up
redo Roll back most recent migration, then apply it again
version Show current migration version
migrate <n> Apply migrations -n|+n
goto <v> Migrate to version v
help Show this help
'-path' defaults to current working directory.
版本控制
将其保存在名为”schema_migrations”的表中。
> select * from schema_migrations;
+------------+
| version |
+------------+
| 1484119339 |
+------------+
1 row in set (0.00 sec)
rubenv/sql-migrate 可以将数据库迁移工作自动化。
数据库处理
-
- MySQL
-
- MSSQL
-
- Oracle Database
-
- PostgreSQL
- SQLite
安装
go get github.com/rubenv/sql-migrate/...
设定
development:
dialect: mysql
datasource: root:@tcp(127.0.0.1:3306)/sqlmigrate_test?parseTime=true
dir: migrations/mysql # ディレクトリを作っておく必要がある
杂感
- 対応RDBMSが多い
迁移文件
没有雏形制作功能。
-- +migrate Up
-- SQL in section 'Up' is executed when this migration is applied
CREATE TABLE `posts` (
`post_id` INT NOT NULL PRIMARY KEY AUTO_INCREMENT,
`author` VARCHAR(255) NOT NULL,
`text` TEXT NOT NULL
) ENGINE = InnoDB;
-- +migrate Down
-- SQL section 'Down' is executed when this migration is rolled back
DROP TABLE `posts`;
Command. (命令)
$ sql-migrate
usage: sql-migrate [--version] [--help] <command> [<args>]
Available commands are:
down Undo a database migration
redo Reapply the last migration
status Show migration status
up Migrates the database to the most recent version available
版本管理
默认情况下,它将保存在名为gorp_migrations的表中。
> select * from gorp_migrations;
+--------------------+---------------------+
| id | applied_at |
+--------------------+---------------------+
| 1_create_posts.sql | 2017-01-11 08:12:18 |
+--------------------+---------------------+
1 row in set (0.00 sec)
埃尔维纳/漫步者
数据库适配
-
- MySQL
-
- PostgreSQL
- SQLite
安装
go get github.com/elwinar/rambler
设定
{
"driver": "mysql",
"protocol": "tcp",
"host": "localhost",
"port": 3306,
"user": "root",
"password": "",
"database": "rambler_test",
"directory": ".",
"table": "migrations"
}
杂感
-
- 設定ファイルがJSONで良い
- ログがとても見にくい
迁移文件
没有创建原型功能。文件将以version_description.sql的命名规则创建。
-- rambler up
CREATE TABLE `posts` (
`post_id` INT NOT NULL PRIMARY KEY AUTO_INCREMENT,
`author` VARCHAR(255) NOT NULL,
`text` TEXT NOT NULL
) ENGINE = InnoDB;
-- rambler down
DROP TABLE `posts`;
指令
$ rambler
11:38 debug loading configuration from%!(EXTRA string=)unable to load configuration from file: open : no such file or directory
NAME:
rambler - Migrate all the things!
USAGE:
rambler [global options] command [command options] [arguments...]
AUTHOR:
Romain Baugue <romain.baugue@elwinar.com>
COMMANDS:
apply apply the next migration
reverse reverse the last migration
help, h Shows a list of commands or help for one command
GLOBAL OPTIONS:
--configuration value, -c value path to the configuration file
--environment value, -e value set the working environment (default: "default")
--debug display debug messages
--help, -h show help
--version, -v print the version
unable to load configuration from file: open : no such file or directory
版本管理
> select * from migrations;
+---------------------+
| migration |
+---------------------+
| 01_create_posts.sql |
+---------------------+
1 row in set (0.00 sec)
闹哪儿/米谷
数据库管理
- MySQL
安装
go get -u github.com/naoina/migu/cmd/migu
杂感
-
- 細かいことを気にしなければ楽ちん
- 融通はきかなそう
迁移文件
我要写一个Go的struct。
package main
type Post struct {
PostId int `migu:"pk,autoincrement"`
Author string
Text string `migu:"size:65535"`
}
移民
以这种方式从struct生成模式并执行给我。
$ migu sync -u root -h 127.0.0.1:3306 migu_test schema.go
--------applying--------
CREATE TABLE `post` (
`post_id` INT NOT NULL PRIMARY KEY AUTO_INCREMENT,
`author` VARCHAR(255) NOT NULL,
`text` MEDIUMTEXT NOT NULL
)
--------done 0.019s--------
命令
$ migu
migu: too few arguments
Usage: migu [OPTIONS] COMMAND [ARG...]
Commands:
sync synchronize the database schema
dump dump the database schema as Go code
Options:
--help Display this help and exit
$ migu sync
migu: too few arguments
Usage: migu sync [OPTIONS] DATABASE [FILE]
Options:
--dry-run Print the results with no changes
-q, --quiet Suppress non-error messages
-u, --user=NAME User for login to database if not current user
-h, --host=HOST Connect to host of database
-u, --user=NAME User for login to database if not current user
-p, --password[=PASS] Password to use when connecting to server.
If password is not given, it's asked from the tty
--help Display this help and exit
With no FILE, or when FILE is -, read standard input.
木匠/dev-cloverlab
对应的数据库
- MySQL
安装
go get github.com/dev-cloverlab/carpenter/cmd/carpenter
混乱的情绪
- 面白い
翻译:模式导出
木工师可以使用特殊工具来进行以下的两种操作。
-
- 実DBのスキーマをJSONへエクスポート
- JSONのスキーマによるマイグレーション
以JSON为基础的模式相当复杂,似乎不适合由人来编写。
导出可以使用如下命令完成。
carpenter --schema goose_test --data-source 'root:@tcp(127.0.0.1:3306)' design
将创建一个名为 tables.json 的文件。
jq "." < tables.json | wc -l
248
$ jq "." < tables.json | head -30
[
{
"TableCatalog": "def",
"TableSchema": "goose_test",
"TableName": "goose_db_version",
"TableType": "BASE TABLE",
"Engine": "InnoDB",
"Version": 10,
"RowFormat": "Compact",
"TableRows": 14,
"AvgRowLength": 1170,
"DataLength": 16384,
"MaxDataLength": 0,
"IndexLength": 16384,
"DataFree": 0,
"AutoIncrement": 17,
"TableCollation": "utf8_general_ci",
"CheckSum": null,
"CreateOptions": "",
"TableComment": "",
"Columns": [
{
"TableCatalog": "def",
"TableSchema": "goose_test",
"TableName": "goose_db_version",
"ColumnName": "id",
"OrdinalPosition": 1,
"ColumnDefault": null,
"Nullable": "NO",
"DataType": "bigint",
移民
如果有JSON,可以根据它生成表格。
carpenter --schema carpenter_test --data-source 'root:@tcp(127.0.0.1:3306)' build --dir .
如果表已经存在并且模式发生了变化,它似乎会进行相应的调整。
命令
$ carpenter
NAME:
carpenter - Carpenter is a tool to manage DB schema and data
USAGE:
carpenter [global options] command [command options] [arguments...]
VERSION:
0.3.1
AUTHOR:
hatajoe <hatanaka@cloverlab.jp>
COMMANDS:
design Export table structure as JSON string
build Build(Migrate) table from specified JSON string
import Import CSV to table
export Export CSV to table
help, h Shows a list of commands or help for one command
GLOBAL OPTIONS:
--verbose, --vv show verbose output (default off)
--dry-run execute as dry-run mode (default off)
--schema value, -s value database name (required)
--data-source value, -d value data source name like '[username[:password]@][tcp[(address:port)]]' (required)
--help, -h show help
--version, -v print the version
似乎还可以使用CSV进行导入/导出。
请参阅
-
- DBマイグレーションツール作った話
- Go製マイグレーションツールの現状確認