This repository shows several key trade-offs to consider while using Docker and Kubernetes with Java EE applications. The repository hosts the demos for this talk or this lab. A video for the talk is here.
The basic Java EE application used throughout is in the javaee folder.
Factors demostrated include:
- Using thin WARs with Docker. The thin-war folder shows how this is done.
- Using uber jars with Docker. The uber-jar folder shows how this is done.
- Using hollow uber-jars with Docker. The hollow-uber-jar folder shows how this is done.
- Administering applications through admin ports exposed out of Docker (as opposed to just deploying applications within Docker images). The exposed-port folder shows how this is done.
- Auto-deploying applications through mounted external Docker volumes on the host (as opposed to deploying applications within Docker images). The external-volume folder shows how this is done.
- How to make application server administration, clustering, autodiscovery and load-balancing work with Docker networking and Kubernetes cluster deployments. The kubernetes-clustering folder shows how this is done.
- How your CI/CD pipeline looks like with Java EE, Docker and Kubernetes. The kubernetes-devops folder shows how this is done.
The demos use Java EE 8, WebSphere Liberty, Azure Kubernetes Service (AKS) and Azure DevOps Pipelines.