Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP

Simple backup of your MySQL database onto Amazon S3.

branch: master

Fetching latest commit…

Octocat-spinner-32-eaf2f5

Cannot retrieve the latest commit at this time

Octocat-spinner-32 bin
Octocat-spinner-32 lib
Octocat-spinner-32 .gitignore
Octocat-spinner-32 CHANGELOG.md
Octocat-spinner-32 Gemfile
Octocat-spinner-32 README.md
Octocat-spinner-32 Rakefile
Octocat-spinner-32 mys3ql.gemspec
README.md

mys3ql = mysql + s3

Simple backup of your MySQL database onto Amazon S3.

Quick start

Install and configure as below.

To perform a full backup:

$ mys3ql full

If you are using MySql's binary logging (see below), back up the binary logs like this:

$ mys3ql incremental

To restore from the latest backup (plus binlogs if present):

$ mys3ql restore

By default mys3ql looks for a configuration file at ~/.mys3ql. You can override this like so:

$ mys3ql [command] -c FILE
$ mys3ql [command] --config=FILE

Installation

First install the gem:

$ gem install mys3ql

Second, create your config file:

mysql:
  # Database to back up
  database:
  # MySql credentials
  user:
  password:
  # Path (with trailing slash) to mysql commands e.g. mysqldump
  bin_path: /usr/local/mysql/bin/
  # If you are using MySql binary logging:
  # Path to the binary logs, should match the log_bin option in your my.cnf.
  # Comment out if you are not using mysql binary logging
  bin_log: /var/lib/mysql/binlog/mysql-bin

s3:
  # S3 credentials
  access_key_id: XXXXXXXXXXXXXXXXXXXX
  secret_access_key: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
  # Bucket in which to store your backups
  bucket: db_backups
  # AWS region your bucket lives in.
  # (I suspect you only need to specify this when your 'location' is in a different region.)
  #region: eu-west-1

If you only have one database to back up on your server, you can put the config file at ~/.mys3ql. Otherwise, tell the mys3ql command where the config file is with the --config=FILE switch.

Binary logging

To use incremental backups you need to enable binary logging by making sure that the MySQL config file (/etc/my.cnf) has the following line in it:

log_bin = /var/db/mysql/binlog/mysql-bin

The MySQL user needs to have the RELOAD and the SUPER privileges. These can be granted with the following SQL commands (which need to be executed as the MySQL root user):

GRANT RELOAD ON *.* TO 'user_name'@'%' IDENTIFIED BY 'password';
GRANT SUPER ON *.* TO 'user_name'@'%' IDENTIFIED BY 'password';

You may need to run mys3ql's incremental backup with special permissions (sudo) depending on the ownership of the binlogs directory.

N.B. the binary logs contain updates to all the databases on the server. This means you can only switch on incremental backups for one database per server, because the logs will be purged each time a database is dumped.

Inspiration

Marc-André Cournoyer's mysql_s3_backup.

To Do

  • tests ;)
  • remove old dump files (s3)
  • (restore from non-latest dump)

Questions, Problems, Feedback

Please use the GitHub issue tracker or email me.

Intellectual property

Copyright 2011 Andy Stewart (boss@airbladesoftware.com).

Released under the MIT licence.

Something went wrong with that request. Please try again.