Deployments
Overview​
The Takeoff Stack is built using Kubernetes and designed to be scalable and fault-tolerant.Our setup is cloud-agnostic and can be deployed with any provider or on-premise. Taking ownership of the infrastructure allows us to provide monitoring and scaling features that are tailored to your needs.
Key Features​
Easy Deployments​
We use a variety of tools to make our deployments as easy as possible.
Dynamic Configuration​
We use Helm to manage our Kubernetes manifests. Helm is a package manager for Kubernetes that allows us to define our deployments in a simple and easy to understand way. This also enables use to dynamically inject configuration into your deployments for customisation purposes.
Continuous Deployment​
We achieve Continuous Deployment/Delivery with ArgoCD. This allows us to verbosely monitor the state of any deployed app and audit its resources. We can set up different sync methods, so you can automatically get new versions or manually approve them. We will consult with you to decide the best method for your use case.
Multiple Environments​
With the use of Helm and ArgoCD we can easily deploy to multiple environments. We can set up applications to match your current deployment flows and provide different sync frequencies. You can set smaller resource allocations for testing and staging environments to save compute.
Self-Healing​
A key feature of Kubernetes is its Liveness, Readiness and Startup Probes, these define the health of the components of the application. It can often be hard to diagnose the cause of user reported issues in a multi-component application. These probes allow us to define meaningful health definitions of individual components and Kubernetes will automatically restart the component if it fails these checks.