Containerizing at scale with Kubernetes
We have already covered how to use containers for building and deploying our ML solutions. The next step is understanding how to orchestrate and manage several containers to deploy and run applications at scale. This is where the open source tool K8s comes in.
K8s is an extremely powerful tool that provides a variety of different functionalities that help us create and manage very scalable containerized applications, including (but not limited to) the following:
- Load Balancing: K8s will manage routing incoming traffic to your containers for you so that the load is split evenly.
- Horizontal Scaling: K8s provides simple interfaces so that you can control the number of container instances you have at any one time, allowing you to scale massively if needed.
- Self Healing: There is built-in management for replacing or rescheduling components that are not passing health checks.
- Automated Rollbacks: K8s stores the history of your...
