Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Makes Amazon S3 backups redonkulous. Redonkulously easy, that is.
Ruby
branch: master

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.
config
examples
lib
tasks
CHANGELOG
MIT-LICENSE
README
init.rb

README

BackupFu
========

The backup_fu plugin makes it redonkulously easy to:

 A) dump your database and/or static files to tar/gzipped or zipped archives, and
 B) upload these archives to a private Amazon S3 bucket for safekeeping

Allows restore of PostgreSQL databases.

It also uses credentials in config/amazon_s3.yml if not supplied in backup_fu.yml

Installation
============

The 'right_aws' gem is required for backup_fu to function properly.  Install with:

  sudo gem install right_aws

Install the plugin with:

  script/plugin install git://github.com/gravelpup/backup_fu.git


Configuration
=============

Run the following to copy over the example backup_fu.yml config file:

  rake backup_fu:setup

This copies the example config file to: config/backup_fu.yml.

Usage
=====

For the list of available rake tasks:

  rake -T backup_fu

Backing up your database:

  rake backup

Dumping your database:

  rake backup_fu:dump

Backing up your static files:

  rake backup_fu:static:backup

Backing up both your database + static files:

  rake backup_fu:all

Restoring from S3:
  BACKUP_FILE=myapp_1999-12-31_12345679_db.tar.gz rake backup_fu:restore
  
Advanced Configuration
======================

See vendor/plugins/backup_fu/config/backup_fu.yml.advanced_example for
the list of advanced configuration options.

Advanced options include:
* specify static path(s) that should be backed up -- i.e. backup your entire 'public/static' directory
* change default dump path from RAILS_ROOT/tmp/backup to whatever
* specify fully-qualified 'mysqldump' path
* disable compression of database dump
* choose between zip or tar/gzip compression
* enable 'nice' with level specification to prevent backup_fu from bogarting your server

Cronjob Installation
====================

Here are some cron job examples.

	# Backup just the database everyday at 1am
	0 1 * * * cd /apps/foo/current; RAILS_ENV=production rake backup > /dev/null

  # Backup db + static @ 2am every 3 days, log the results to ~/backup.log (verbosity should be turned on if logging results)
  0 2 1-31/3  * * cd /u/apps/shanti.railsblog/current; RAILS_ENV=production rake backup_fu:all >> ~/backup.log


Debugging
=========

--- Enabling Verbosity

If you are experiencing any difficulties, the first thing you should do is enable verbosity by
dropping this into config/backup_fu.yml:

  verbose: true
	
--- Mysqldump Issues

If your 'mysqldump' command is not in your path, you will need to specify it explicitly.

To see if mysqldump is in your path, execute:

  which mysqldump

If you see output like "/usr/bin/which: no mysqldump in (...)" then you will need to specify the path manually.

Use 'locate mysqldump' or a similar tool to find the full path to your mysqldump utility.

Place an entry like the following in your config/backup_fu.yml file:

  mysqldump_path: /usr/local/mysql-standard-5.0.27-linux-i686/bin/mysqldump

--- Database Connection Issues

If you are seeing an error when running 'rake backup' like:

  mysqldump: Got error: 2002: Can't connect to local MySQL server ...

Make sure you are specifying the RAILS_ENV for the target environment.  i.e. for production:

  RAILS_ENV=production rake backup
  or
  rake backup RAILS_ENV=production

--- Connection reset by peer

When backing up, if you receive an error like:

  rake aborted!
  Connection reset by peer

Chances are this is because your backup is huuge.  There is currently no great solution for
this problem.

On some systems, I have backed up 4GB+ files without a hitch.  On other machines, an 80mb
backup was choking on the S3 upload.  After 3 attempts it went through.

Patching in some kind of email notification system on failure would probably be nice.

Patches welcome =)

--- Tiny Static file .tar.gz Archive (static files not actually getting archived)

This may result if you are using a symlink for your static dir, such as:

  public/static -> /shared/apps/foo/static

The solution to this is to specify the absolute static path in config/backup_fu.yml:
 
  static_paths: /shared/apps/foo/static



Copyright (c) 2009, released under the MIT license
Something went wrong with that request. Please try again.