⇩ Markdown

blog-post-getting-rails-on-kubernetes

https://engineering.adwerx.com/rails-on-kubernetes-8cd4940eacbe#.s3ufpxe0w

link not tracked

Kubernetes master may also be a worker node and have pods scheduled to it.

Some commands

kubectl get pods

kubectl describe pod [pod name] (also shows last event)

kubectl create -f ./some/path

kubectl apply -f ./some/path (to make changes)

kubectl port-forward [pod name] 3606:3606 (get access to it on your local machine)

kubectl delete job deploy-tasks

kubectl describe job deploy-tasks

kubectl attach (TODO - what is this?)

^some-commands

livenessProbe and readynessProbe for health checks

In a deployment, nodeSelector will give you an affinity for a specific node-pool.

There is a load balancer service.

HorizontalPodAutoscaler

Within that file, there is an apiVersion: extensions/v1beta1

He creates a template that he pipes into kubectl (adds .tmpl suffix to the file). For that he uses envsubst in the gettext package.

In his script he uses jsonpath.

Collecting metrics

dd-agent kubernetes-docs-deamon-set spec - sends health of every node, pod, and kube-system service to Datadog

^dd-agent-and-daemon-set

References

link not tracked


kubernetes-glossary


tag--consumption-notes

tag--project-kubernetes

tag--pub-to-kubnotes