heroku-db-backup-s3

by andycloke

GitHub Readme.md

Heroku Buildpack: heroku-db-backup-s3

  • A fork of devsenexx/heroku-db-backup-s3. This stores in .dump format not plain text, in order to make restoring the db easier. It also uses the DATABASE_URL environment variable directly, so you dont have to duplicate it.

Capture Postgress DB in Heroku and copy it to s3 bucket. Buildpack contains AWS CLI.

Installation

Add buildpack to your Heroku app

heroku buildpacks:add https://github.com/andycloke/heroku-db-backup-s3 --app <your_app>

Buildpacks are scripts that are run when your app is deployed.

Configure environment variables

heroku config:add AWS_ACCESS_KEY_ID=someaccesskey --app <your_app>
heroku config:add AWS_SECRET_ACCESS_KEY=supermegasecret --app <your_app>
heroku config:add AWS_DEFAULT_REGION=eu-central-1 --app <your_app>
heroku config:add S3_BUCKET_PATH=your-bucket --app <your_app>
heroku config:add HEROKU_TOOLBELT_APP=<your_app> --app <your_app>
heroku config:add HEROKU_TOOLBELT_API_EMAIL=sss --app <your_app>
heroku config:add HEROKU_TOOLBELT_API_PASSWORD=ddd --app <your_app>

### Scheduler
Add addon scheduler to your app.

heroku addons:create scheduler --app <your_app>

Create scheduler.

heroku addons:open scheduler --app <your_app>

Now in browser Add new Job. Paste next line: bash /app/vendor/backup.sh -db <somedbname> and configure FREQUENCY. Paramenter db is used for naming convention when we create backups. We don't use it for dumping database with the same name.

Doesn't work?

In case if scheduler doesn't run your task, check logs using this e.g.:


heroku logs -t --app <your_app> | grep 'backup.sh'
heroku logs --ps scheduler.x --app <you_app>