Backup To Amazon S3

This script can be used to back up virtually any LAMP application to Amazon S3. It performs a mysqldump on one or more databases, and backs up files from one or more folders. It is inspired by the COPIOUS Magento Backup script at https://github.com/copious/magento_backup

Prerequisites

PHP 5.3 or later is required to run this backup script. Namespaces are used for the core classes, and composer ships as a Phar Archive.

The prerequisite libraries for this project are installed via composer. Once you've cloned this git repository, use composer to install them:

$ php bin/composer.phar install

Running Backups

To use it, drop your authentication details into a backup.yml file as follows:

name: myproj  # Used to name the tarball we upload to S3
amazon:
  bucket: my_bucket_name   # Must be created manually
  secret_access_key: 'secret key here'
  access_key_id: access_key_here
database:  # Backup a database with mysqldump
  -        # Repeat this block if you need to back up multiple databases.
    name: main   # Used to identify the database when creating filenames,
    username: mysql_username
    password: 'mysql_password'
    hostname: localhost
    database: name_of_database
    touch: /wwwroot/myproj/maintenance.flag  # (optional) Create a file while the database backup is performed and delete it afterwards.  Can be used to put magento into maintenance mode.
files: # Backup a foilder including all files and subfolders.
  -    # Repeat this block if you want to back up multiple filder.
    name: assets      # Used to identify this path when creating backjup folder,
    path: /wwwroot/myproj/media

Once it's configured, run the backup:

	$ php backup.php

In the grand Unix tradition, the script will produce no output on success, If an error occurs, Error messages will be displayed and the script will return a non 0 status.

If multiple .yml configuration files exist, you can select which one to use from the command line:

	$ php backup.php --config otherconfig.yml

The script utilises the multipart upload capabilities of the AWS SDK to overcome the 5Gb POST size limit.

Restoring

To restore from a backup, you'll first need to download from Amazon S3 the backup tarball you intend to restore. The S3 web client at http://aws.amazon.com/ is the easiest way to do this. Then you'll need to extract the archive using tar and gzip:

  $ tar zxvf my-backup.tgz

This will create a directory containing the assets, and database dump as originally backed up..

Encrypted Backups

You can GPG encrypt backups before being exporting to Amazon S3 by adding the following to your backup.yml

gpg:
  encryption_key: "ABCD1234"  # Supply key id here.

The backup script will then encrypt the backup using the supplied GPG public key. Ensure you have a copy of the corresponding Secret Key and Passphrase stored in a safe place. If you loose the secret key your backups will be irrecoverable.

Make sure the public key is trusted so that GPG runs without prompting with the following message:

It is NOT certain that the key belongs to the person named
in the user ID.  If you *really* know what you are doing,
you may answer the next question with yes.

Use this key anyway? (y/N)

If you see this, either sign the key or mark it as "ultimate" trust. To do the latter run gpg --edit-key <key id> trust, select option 5 "I trust ultimately" and then type quit to exit.

Encrypted Restores

To restore a backup that was encrypted, you must have a copy of the secret key from the keypair used to make the backup. If the secret key is not available, the backup will be irrecoverable. To check if you have a copy of the required secret key on your GPG keyring, the command gpg --list-secret-keys can be used. This will display a list of the secret keys available on your keyring. If you don't have the required key on your keyring, obtain the key and place it in a file on your filesystem. Then run gpg --import <filename> to import it. After importing, use the "shred" command to securely delete the temporary file.

To actually restore the backup, fetch the .gpg file from Amazon S3, and run the command gpg -d <filename>. You'll be prompted to enter the passphrase for the secret key, then the file will be decrypted. Once decrypted, you can follow the normal restoration instructions above.