We’ve made several improvements to our Kubernetes stacks, allowing us to deploy in different AWS Regions (eg. us-east-1) and allowing more dynamic usage of the Availability Zones in those regions.
Before we were moslty focussed on the eu-west-1 region, since that’s also where our customers mostly situated. However the need to deploy in other Regions arised, triggering the following changes:
- Publish our AMIs in the
us-east-1andeu-central-1Regions, next toeu-west-1. More Regions can easily be added on a per-need basis via our automated pipelines. - Updated our automation code and pipelines to support different Regions.
- We now create dedicated subnets for running EKS nodes, instead of using the default
private_appsubnets generated by our VPC module. - Allow specifying which (and how much) Availability Zones to use for the EKS node groups. For example this allows excluding AZs with limited capacity, or using a specific AZ only for a node group.