One Rust and one golang microservice that are deployed to microk8s. There is a cron job defined on the cluster that runs every minute and calls a random enpoint on a microservice. Each microservice has an endpoint /die
that will cause the pod to die forcing the HorizontalPodAutoscaler to create a new pod.
Each microservice has a /metrics
endpoint that is scraped by prometheus. The prometheus server is deployed to the cluster and is configured to scrape the microservices.
./setup-infra.sh
microk8s kubectl describe secret -n kube-system microk8s-dashboard-token
./build-and-deploy.sh
- microk8s
- docker
- rust
- golang
- java
- javascript
- kafka
- https://spring.io/projects/spring-boot
- https://go.dev/doc/
- https://www.rust-lang.org/learn
- https://microk8s.io/
-
publish metrics of service kill calls (prometheus?/kafka?) => kafka
-
create a reactive be4fe(?) that handles websocket connections to push messages to a UI (react?) that shows when a service is killed