ridoku is intended to be a set of scripts for replacing a Heroku work flow with AWS OpsWorks. It requires some manual configuration at the moment in AWS OpsWorks and the IAM Control Panel.

You must add a user or add full permission to OpsWorks to your existing user.
The easiest way to do this is by following this guide. Once you have this completed, you should be able to use ridoku to manage certain aspects of your app deployment.

Currently, ridoku only works with the Rails Application Stack (rather, its only been tested on the stack, and several layer-actions specify 'rails-app' type, so your mileage may vary).

Also, when developing the scripts, I was using a Rails app with a custom PostgreSQL, so the use of the OpsWorks MySQL layer has not been tested, but probably won't be affected (it will probably only limit the use of the ridoku db command set for application management).

Once the Stack has been created and instances added using the OpsWorks GUI, you should be able to start using ridoku to make edits to your database and environment information, as well as running recipes and commands on the stack.

Sadly, at this point, the OpsWorks GUI is still required (stack config, layers, etc are manual atm).

Currently, the Ridoku custom cookbooks are also required to ensure that the environment is the same as is expected by a Heroku application.

If you have a Stack that is already configured to use ridoku, this section gives you a quick run-down of commands necessary for Application management.

(rid can also be used as an alias for ridoku)

Each command below expects you to have run:

$ ridoku --set-stack YourStack
$ ridoku --set-app YourApp
$ ridoku --set-backup-bucket YourBackupBucket

The switches --app app-name and --stack stack-name can be used in any given commandline to override defaults.

$ ridoku deploy

Deploys the application to all instances. Note that HEAD is used for the repository branch associated with this app. (This is currently only configurable in the OpsWorks console)

$ ridoku deploy:rollback

Rollback the application on all instances.

These commands only work if you are using ridoku to manage databases.

$ ridoku backup:capture

Captures the current applications database and stores it to S3.

$ ridoku backup:list
$ ridoku backup:capture
$ ridoku backup:restore <backup name>

Shows all existing database backups for the specified application, captures a backup (safety first!), then restores the specified database backup.

compare to heroku config

All changes to the environment require an application deploy to take effect. The Revision provider is used in the ridoku deployment cookbooks. As a result, multiple deploy commands can be issued in a row without depleting the rollback capability (which is limited to 5 total rollbacks).

$ ridoku env

Displays the current applications runtime environment configuration.

$ ridoku env:set KEY:value KEY2:value2

Sets or updates the specified key/value pairs.

$ ridoku env:remove KEY

Removes the specified key/value pair.

Ridoku, 0.0.8

usage: ridoku [OPTIONS] command [command options]

TODO

TODO

TODO

TODO

TODO

TODO

TODO

TODO

TODO

TODO

TODO

TODO

TODO

In order to get ridoku configured with your OpsWorks account, Ridoku must collect pertinent required info. The wizard can be run at any time after the first with the command line option of --wizard.

Path to the SSH key to be used for git repositories (cook books, apps, etc). It is recommended that this be generated separately from your personal SSH keys so that they can be revoked effecting other logins.

If a valid service_role_arn cannot be found, Ridoku will attempt to generate one for you. If you've already used OpsWorks, Ridoku should be able to find the necessary Roles for you.

If a valid instance_role_arn cannot be found, Ridoku will attempt to generate one for you. If you've already used OpsWorks, Ridoku should be able to find the necessary Roles for you.

Amazon OpsWorks similarly to Heroku, but, because you manage all the resources, you'll have to provide a bit more information than you do to Heroku in order for this commandline utility to assist.

The technology stack to use for a particular type of application.

Heroku probably has a similar structure internally, because they allow you to use any number of program Stacks (Rails, PHP, Python, Go, etc). The difference is that now in OpsWorks you control the stack environment, where on Heroku you did not.

If you have a stack configured, you can view pertinent information using

$ ridoku list:stacks

This will display the stacks that are currently associated with your AWS account. To

To set the specific stack to use:

$ ridoku --set-stack <stackname>

To set a single run stack or override the default:

$ ridoku --stack <stackname> --app <appname> command…

The actual application which runs on the technology stack.

This is what you have control over on Heroku. You can customize the app domains, database information, environment, etc, on a per-application basis. The same goes for OpsWorks.

To set the default app to use:

$ ridoku --set-app <stackname>

To set a specific run app or override the default:

$ ridoku --stack <stackname> --app <appname> command…

I would like to get this to the point of a fully functional Heroku replacement. Adding a standard Rails stack using the standard LB -> Web Server*N <-> DB stack layout should be fairly easily accomplished.

If you have any issues when attempting to use this toolchain, please feel free to submit a pull request.