Mysql_Framework

Welcome to Mysql_Framework, this is a lightweight framework that provides managers to help with interacting with mysql.

Add this line to your application's Gemfile:

gem 'mysql_framework'

• MYSQL_HOST - MySQL Host
• MYSQL_PORT - MySQL Port
• MYSQL_DATABASE - MySQL database name
• MYSQL_USERNAME - MySQL username
• MYSQL_PASSWORD - MySQL password

• MYSQL_READ_TIMEOUT - how long before connections time out when reading information from the DB (default: 30 seconds)
• MYSQL_WRITE_TIMEOUT - how long before connections time out when writing information to the DB (default: 10 seconds)

• MYSQL_START_POOL_SIZE - how many connections should be created by default (default: 1)
• MYSQL_MAX_POOL_SIZE - how many connections should the pool be allowed to grow to (default: 5)

• MYSQL_MIGRATION_TABLE - the name of the table that holds a record of applied migrations (default: migration_script_history)
• MYSQL_MIGRATION_LOCK_TTL - how long the tables should be locked for whilst performing migrations (default: 2000 / 2 seconds)
• MYSQL_MIGRATION_LOCK_MAX_ATTEMPTS - how many times the lock manager should attempt to acquire the lock before failing (default: 300)
• MYSQL_MIGRATION_LOCK_RETRY_DELAY_S - how long the lock manager should sleep between lock request attempts (default: 1 second)
• REDIS_URL - The URL for redis - used for managing locks for DB migrations

• MYSQL_PARTITIONS - if a table is partitioned, how many partitions should be created (default: 500)

Migration scripts need to be in the following format:

class CreateDemoTable < MysqlFramework::Scripts::Base
  def initialize
    @identifier = 201806021520 # 15:20 02/06/2018
  end

  def apply(client)
   client.query(<<~SQL)
      CREATE TABLE IF NOT EXISTS `#{table_name}` (
        `id` CHAR(36) NOT NULL,
        `name` VARCHAR(255) NULL,
        `created_at` DATETIME NOT NULL,
        `updated_at` DATETIME NOT NULL,
        PRIMARY KEY (`id`)
      )
    SQL
  end

  def rollback(client)
    client.query(<<~SQL)
      DROP TABLE IF EXISTS `#{table_name}`
    SQL
  end

  def tags
    [table_name]
  end

  private

  def table_name
    DemoTable::NAME
  end
end

The initialize method should set the @identifier value, which should be a timestamp:

@identifier = 201806021520 # 15:20 02/06/2018

Make sure @identifier is an integer too, otherwise MysqlFramework::Scripts::Manager may struggle to determine which are your pending migrations.

The apply method should action the migration. An instance of Mysql2::Client is available as client to use.

The rollback method should action the migration. An instance of Mysql2::Client is available as client to use.

Tags are used for when we want to specify which migrations to run based on a tag. This is useful for tests where you don't need to run all migrations to assert something is working or not.

Use the MysqlFramework::Scripts::Manager#execute method to run all pending migrations.

Used to register tables. This is used as part of the all_tables method in the script manager for awareness of tables to drop.

class DemoTable
  extend MysqlFramework::Scripts::Table

  NAME = 'demo'

  register_table NAME
end

The connector deals with the connection pooling of MySQL2::Client instances, providing a wrapper for queries and transactions.

connector = MysqlFramework::Connector.new
connector.setup
connector.query(<<~SQL)
  SELECT * FROM gems
SQL

Options can be provided to override the defaults as follows:

options = {
  host: ENV.fetch('MYSQL_HOST'),
  port: ENV.fetch('MYSQL_PORT'),
  database: ENV.fetch('MYSQL_DATABASE'),
  username: ENV.fetch('MYSQL_USERNAME'),
  password: ENV.fetch('MYSQL_PASSWORD'),
  reconnect: true
}
MysqlFramework::Connector.new(options)

Sets up the connection pooling. Creates ENV['MYSQL_START_POOL_SIZE'] Mysql2::Client instances up front. This is provided as a separate method to allow for use within process forking where connections would need to be created after forking the process.

connector.setup

Closes all the Mysql2::Client connections and removes the connection pool. Intended as a clean-up method to be used on process fork shutdown.

connector.dispose

Check out a client from the connection pool. Will create new Mysql2::Client instances up-to ENV['MYSQL_MAX_POOL_SIZE'] times if no idle connections are available.

client = connector.check_out

Check in a client to the connection pool

client = connector.check_out
# ...
connector.check_in(client)

Called with a block. The method checks out a client from the pool and yields it to the block. Finally it ensures that the client is always checked back into the pool.

connector.with_client do |client|
  client.query(<<~SQL)
    SELECT * FROM gems
  SQL
end

It can optionally accept an existing client to avoid starting new connections in the middle of a transaction. This can be used to ensure that a series of queries are wrapped by the same transaction.

connector.with_client(existing_client) do |client|
  client.query(<<~SQL)
    SELECT * FROM gems
  SQL
end

This method is called when executing a prepared statement where value substitution is required:

insert = MysqlFramework::SqlQuery.new.insert(gems)
  .into(gems[:id],gems[:name],gems[:author],gems[:created_at],gems[:updated_at])
  .values(SecureRandom.uuid,'mysql_framework','sage',Time.now,Time.now)

connector.execute(insert)

It can optionally accept an existing client to avoid checking out a new client.

insert = MysqlFramework::SqlQuery.new.insert(gems)
  .into(gems[:id],gems[:name],gems[:author],gems[:created_at],gems[:updated_at])
  .values(SecureRandom.uuid,'mysql_framework','sage',Time.now,Time.now)

connector.execute(insert, existing_client)

This method is called to execute a query without having to worry about obtaining a client

connector.query(<<~SQL)
  SELECT * FROM versions
SQL

It can optionally accept an existing client to avoid checking out a new client.

connector.query(<<~SQL, existing_client)
  SELECT * FROM versions
SQL

This method requires a block and yields a client obtained from the pool. It wraps the yield in a BEGIN and COMMIT query. If an exception is raised then it will submit a ROLLBACK query and re-raise the exception.

insert = MysqlFramework::SqlQuery.new.insert(gems)
  .into(gems[:id],gems[:name],gems[:author],gems[:created_at],gems[:updated_at])
  .values(SecureRandom.uuid,'mysql_framework','sage',Time.now,Time.now)

connector.transaction do |client|
  client.query(insert)
end

The default options used to initialise MySQL2::Client instances:

{
  host: ENV.fetch('MYSQL_HOST'),
  port: ENV.fetch('MYSQL_PORT'),
  database: ENV.fetch('MYSQL_DATABASE'),
  username: ENV.fetch('MYSQL_USERNAME'),
  password: ENV.fetch('MYSQL_PASSWORD'),
  reconnect: true
}

A representation of a MySQL Condition for a column. Created automatically by SqlColumn

# eq condition
SqlCondition.new(column: 'name', comparison: '=', value: 'mysql_framework')

A representation of a MySQL column within a table. Created automatically by SqlTable.

SqlCondition.new(table: 'gems', column: 'name')

A representation of a MySQL Query.

gems = MysqlFramework::SqlTable.new('gems')
guid = SecureRandom.uuid

# Insert Query
insert = MysqlFramework::SqlQuery.new.insert(gems)
  .into(gems[:id],gems[:name],gems[:author],gems[:created_at],gems[:updated_at])
  .values(guid,'mysql_framework','sage',Time.now,Time.now)

# Update Query
update = MysqlFramework::SqlQuery.new.update(gems)
  .set(updated_at: Time.now)
  .where(gems[:id].eq(guid))

# Delete Query
delete = MysqlFramework::SqlQuery.new.delete
  .from(gems)
  .where(gems[:id].eq(guid))

# Bulk Values Query
bulk_insert = MysqlFramework::SqlQuery.new.insert(gems)
  .into(gems[:id],gems[:name],gems[:author],gems[:created_at],gems[:updated_at])
  .bulk_values([[guid,'mysql_framework','sage',Time.now,Time.now], [guid,'mysql_framework','sage',Time.now,Time.now]])

# Bulk On Duplicate Query
bulk_upsert = MysqlFramework::SqlQuery.new.insert(gems)
  .into(gems[:id],gems[:name],gems[:author],gems[:created_at],gems[:updated_at])
  .bulk_values([[guid,'mysql_framework','sage',Time.now,Time.now], [guid,'mysql_framework','sage',Time.now,Time.now]])
  .on_duplicate(gems[:id] => nil,gems[:name] => nil,gems[:author] => nil,gems[:created_at] => nil,gems[:updated_at] => nil)

A representation of a MySQL table.

MysqlFramework::SqlTable.new('gems')

As a default, MysqlFramework will log to STDOUT. You can provide your own logger using the logger= method:

MysqlFramework.logger = Logger.new('development.log')

Bug reports and pull requests are welcome on GitHub at https://github.com/sage/mysql_framework. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the Contributor Covenant code of conduct.

A compose file is provided for running specs.

docker-compose up -d
docker-compose exec test-runner bash
# Once the shell opens in the container
bundle

bundle exec rspec

Exit out of the shell when finished.

docker-compose down

This gem is available as open source under the terms of the MIT licence.

Copyright (c) 2018 Sage Group Plc. All rights reserved.