Automate Heroku database backups

Backup and download database backup

#!/bin/bash
#
# Script to create and download a database backup on Heroku
# Written by Chris Stefano (https://github.com/virtualstaticvoid)
#
# See http://devcenter.heroku.com/articles/pgbackups for details
#

appname=$1
filename=${appname}_db_$(date +%Y%m%d)_$(date +%H%M).pg_backup

# list backup
# heroku pgbackups --app $appname

# may need to delete old backups:
# heroku pgbackups:destroy XXXX --app $appname

# create a backup
heroku pgbackups:capture --app $appname

# get the temp URL for it (AWS S3)
backup_url=`heroku pgbackups:url --app $appname`

# download it locally
curl -o $filename $backup_url

Delete second most recent backup

#!/bin/bash

# Script to "clean up" old database backups on Heroku
# Written by Chris Stefano (https://github.com/virtualstaticvoid)
#
# See http://devcenter.heroku.com/articles/pgbackups for details
#
# NOTE: Heroku seems to only allow 2 backups at any given time, presumable on the free tier?, 
#
# !!! USE AT YOUR OWN RISK!!!
#

appname=$1

# get the backup line-items from the `pgbackups` command
backup=`heroku pgbackups --app $appname | grep ^b[0-9]*`

# extract the name of it
backup_name=${backup:0:4}

# BE GONE!!!
heroku pgbackups:destroy $backup_name --app $appname

Hi! Thanks for this. :)

I took the liberty of forking and making a few tweaks of my own; you might be interested:

• used --expire option, removing need for second script (I think)
• changed naming of downloaded file (personal preference), plus used .dump extension
• symlinked to latest dump (I always like having a quick way of accessing the latest dump)

https://gist.github.com/tiredpixel/4774216