Personal repository revolving around kubernetes (k8s).
Kubernetes was never truly designed to be deployed on-premise, but thankfully there are third party components that mimic and replace the need for dedicated cloud components. My on-premise Kuberrnetes deployment configuration is detailed here.
What is Kubernetes?
With my own interpretation...
-
Opensourced "borg"
-
Handling global scale services with 100% uptime
-
Scaling services to handle millions of transactions in different regions of the world
-
HPC (High Performance computing)
-
Iterating over docker and docker compose
-
Other applications?
Repository Structure
.
├── go-endpoint # Standalone REST API test application
├── go-endpoint-mongodb # Basic REST API that interacts with mongoDB
│ └── go-endpoint-mongodb.yaml # Kubernetes deployment config
│ └── test-request.http # Predefined requests to test endpoints
├── mongo # Kubernetes MongoDB and Mongo-express deployment configurations
│ └── docker-compose.yml # For local device docker testing
└── reference_yaml # Place to store yaml configs for reference
├── MyKubernetesConfiguration.md
└── README.mdAcknowledgement section covers any extra knowledge, articles and random thoughts.
- k8s-playground
- My Configured Kubernetes Overview
- Basic Kubernetes feature testing
- Acknowledgements
- Appendix
- Deploy a stacked HA Kubernetes topology cluster via kubeadm
- Create custom manifest to deploy custom
go-endpoint-mongodbpods to interact with mongoDB - Conduct basic stress and load tests to see how kubernete's features work
- Document the knowledge and configurations in a self-hosted environment for future reference
- Load and Stress testing to understand kubernete's automated system
- Prometheus configuration monitoring
- Configuring Storage Orchestration
- Persistent Volume Claim (PVC) --> Persistent Volume (PV)
- Proper Nginx Ingress controller and TLS/SSL termination
- Automated CI/CD setup with Jenkins
To show case what my current configuration is capable of
Image
Kill Pods
(gif)
Kill Nodes
(gif)
Kill a Worker Node
(gif)
Kill a Control Plane
(gif)
Kill a Control Plane and Worker Node
(gif)
As kubeadm is just a tool to create a minimum viable cluster. More in-depth configuration can be done as shown in kelseyhightower and mmumshad repository.
- minikube
- kind
- k3s
- microk8s
-
persistent storage
-
namespaces
-
rbac
-
containerd
-
Flannel (CNI)
-
CoreDNS
-
Traefik (Ingress)
-
Klipper-lb (Service LB)
-
Embedded network policy controller
-
Embedded local-path-provisioner
-
Host utilities (iptables, socat, etc)
Explaining the purpose of kubernetes https://dev.to/thenjdevopsguy/what-problem-is-kubernetes-actually-trying-to-solve-3g1n
Setting up kubernetes in proxmox https://www.learnlinux.tv/how-to-build-an-awesome-kubernetes-cluster-using-proxmox-virtual-environment/
Article regarding high availability kubernetes using kubeadm https://medium.com/velotio-perspectives/demystifying-high-availability-in-kubernetes-using-kubeadm-3d83ed8c458b
How Kube-vip replaces both haproxy and keepalived to a degree https://inductor.medium.com/say-good-bye-to-haproxy-and-keepalived-with-kube-vip-on-your-ha-k8s-control-plane-bb7237eca9fc
Explanation of kube-vip https://thebsdbox.co.uk/2020/01/02/Designing-Building-HA-bare-metal-Kubernetes-cluster/#Networking-load-balancing
Explanation of multi-cluster kubernetes https://traefik.io/glossary/understanding-multi-cluster-kubernetes/
Examples of multi-cluster kubernetes configuration https://www.kubecost.com/kubernetes-multi-cloud/kubernetes-multi-cluster/
Kubernetes for Data Engineering https://medium.com/@vladimir.prus/kubernetes-for-data-engineering-feba247f7585