Roost - Environment As a Service

Owned by Harish Agrawal
Last updated: Aug 06, 2022
2 min read

Using Roost, you can configure once and enjoy automated preview environments. These environments provide you a number of benefits:

  • Ephemeral environment created real-time with your current code changes and ready to be used by all.

  • Immediately accessible as well as shareable across different stake-holders

  • Certification of your changes against the existing test-suite.

  • Identifies changes that are breaking and prevents any production issues OR time lost towards manual QA/testing.

  • Controlled cloud cost as environment auto expires after a certain time.

  • Central dashboard for all your team members to access and view all the preview environments created along with the associated change.

  • Capability to control/limit the number of preview environments, discard events that are yet to be processed (this role is limited to DevOps and Admin members)

  • On-demand ability to view logs, recreate any previous environment or even destroy any environment forever.

  • Ability to apply ad-hoc change and conduct what-if analysis on the ephemeral environments as well.

And finally, Configure just once and reap benefits forever

Simple onboarding steps for any application

Code Repository

  • Provide a GitHub token (GitLab and Bitbucket are in the roadmap)

  • Select Repository and branch (allows multiple repo)

  • Choose Git events that trigger “Roost Environment”

    • Create PR

    • Merge PR

    • Create Tag

Infrastructure Steps

  • Specify Terraform Script/Repo for creating infrastructure resources outside of the cluster

  • AWS CloudFormation support is in the roadmap

  • You can specify optionally Terraform destroy steps as well for execution when the environment is shutdown

Build Steps

  • Specify Build tool for each repo

  • Choose the Build tool and build order (if it matters)

    • Jenkins (Provide token to detect job-name, build status)

    • Custom tools like Make, Maven, Apache Ant, Docker Build, BuildPacks

Cluster Configuration

  • Go with the default cluster configuration (default for organisation)

  • Or specify necessary overrides along with cloud credentials

  • Support for on-premise and existing cluster is in the roadmap

Deploy Steps

  • Choose the deploy tool like Kubernetes, Helm and Skaffold.

  • Provide the necessary application arguments to the above tool

Notification Steps

  • Choose Slack or Email for notification once the release environment is setup

  • Provide necessary authorisation token

Optional Test Suite integration

  • Support for automated testing and certification of change

  • Framework such as artillery and cypress are already supported.

  • Postman API, LambadaTest are in the roadmap