AWS CodeDeploy

At Kloud.io we use Node.js for our backend Microservices and Angular 2 for the frontend. We also use React for a couple of our internal applications. Writing services in Node.js in TypeScript improved developer productivity and we could capture bugs way before they can occur in the production. The use of Angular 2 in our production environment reduced the time to release any new features. At the same time, we are also exploring React by using it in our internal tools. So far we enjoyed what React has to offer. We are an enterprise SAAS product and also offer an on-premise or hybrid cloud version of #kloudio. We heavily use Docker for shipping our on-premise version. We also use Docker internally for automated testing. Using Docker reduced the install time errors in customer environments. Our cloud version is deployed in #AWS. We use AWS CodePipeline and AWS CodeDeploy for our CI/CD. We also use AWS Lambda for automation jobs.

Today we added AWS CodeDeploy to our stack. The easiest way to deploy our code towards a big Autoscaling group of Amazon EC2 instances. The only problem we found on our way was that when using AWS Elastic Load Balancing (ELB) and AWS CodeDeploy , you can only select one ELB Target Group when creating a deployment. While we actually use 5 towards each Amazon EC2 instance because one deployed Node.js code source actually has the data to server 5 vhosts. Luckily we can let the bash script used in AWS CodeDeploy tell the ELB to stop sending it traffic to all Target groups during the in-place update.

The recent move of our CI/CD tooling to AWS CodeBuild / AWS CodeDeploy (with GitHub ) as well as moving to Amazon EC2 Container Service / AWS Lambda for our deployment architecture for most of our services has helped us significantly reduce our deployment times while improving both feature velocity and overall reliability. In one extreme case, we got one service down from 90 minutes to a very reasonable 15 minutes. Container-based build and deployments have made so many things simpler and easier and the integration between the tools has been helpful. There is still some work to do on our service mesh & API proxy approach to further simplify our environment.