5 Popular Deployment Strategies in Cloud-Computing

Canary, A/B Testing, Ramped, Blue/Green, Recreate…

Ali Zeynalli

--

Photo by Kimon Maritz on Unsplash

In Cloud-Computing there are plenty of ways to construct the new release deployment process. The correct strategy must be chosen accurately to have a stable application while new release rollout. In tools like Kuberenetes or Openshift you can preconfigure the exact deployment strategy easily. Without further due, let us have a look at some well-established deployment strategies.

Canary…

In this type of deployment strategy you deploy the new version of the application to small group of users in order to get feedback before heading to the big release. The advantages of Canary Deployment are especially error monitoring, easiness of deployment due to small traffic, fast rollback with no big consequences. On the other side, Canary Deployment brings some overhead in configuration management since you need to distribute server requests accurately between new and old version. In the language of different orchestration tools, you would need to have two replicasets (one for old version pods, one for new version pods) and small number of pods with new version that is directed from loadbalancer when test group tries to reach the app.

A/B Testing…

--

--

Ali Zeynalli

IT-Architect | BMW Group | M.Sc. Technical University of Berlin