In the previous steps, we learned about Octopus Deploy and how it can support your continuous delivery. In this learning step, we will see how Octopus supports multiple advanced deployment patterns.

Rolling Deployments

Rolling deployments allows for gradual deployment: deploying one machine at a time in sequence with pauses in between. This allows you to reduce the risk of deployment through earlier detection of issues in a new release, such as an increase in regular usage patterns (like CPU above the baseline, or slow database accesses). This enables you to stop a deployment before further damage occurs or to perform Canary Releases.

Blue/Green Deployments

A blue/green deployment is a deployment pattern that reduces downtime and risk by having two identical production environments (blue and green). At a given time only one of the set is serving live production traffic. You deploy to the one that is inactive, and when the deployment is finished and eventually validated, you switch them.

If major errors are found, it is easy to roll back by reversing the switch.

Immutable Infrastructure

You can use an immutable infrastructure philosophy by provisioning environments as part of the deployment to make sure that your deployments are fully recreated from scratch every time you deploy an application.

In the next step, we will investigate Platforms, Web Portal, Multitenancy and Packages, but before you move on, please participate in the discussion.

Join the discussion

Octopus Deploy is great when it comes to supporting continuous delivery. Discuss the different aspects of Rolling Deployments and Blue/Green Deployments and give examples of of how these are used.

Use the discussion section below and let us know your thoughts. Try to respond to at least one other post and once you’re happy with your contribution, click the Mark as complete button to check the step off, then you can move to the next step.