Cursus 28
Big Data architectures are evolving with new needs, such as on-demand scalability or application deployments in hybrid environments (mix of multiple public clouds and on-premises environments). Kubernetes is one of the technologies behind this transformation. Kubernetes (K8s) is an open-source system for automating the deployment, scaling and management of containerized applications. The course leads to the understanding, installation, management of this very powerful infrastructure.
The training is aimed at a technical population (IT staff, IS architects) who have solid experience in infrastructure management in the IS and who have a devops experience.
It’s the ideal training to understand the architectures of tomorrow, whether they are Big Data or not.
Program
Day 1
Introduction
The Docke environment
- Principles and architecture
- Creation of image Docker
- Compose & Swarm
Concepts
- Principles and architecture
- Cluster Relationships
- Terminology
- pods, master, node, label, service, replication controller, tools (kubeadm, kubelet, kubectl …)
Pod
- Presentation
- Creation and operation of a Pod
Workloads
- Deployment
- DaemonSets
- Jobs
Installation and configuration
- Become familiar with kubctl
- Create a deployment, a replica set, a pod
How to administer your cluster
Practical work
Day 2
Deployment strategy
- The different strategies
- Rolling update
Expose its applications
- Communication between pods
- Expose a service
Configure your application
- Via files
- Via environment variables
Case of statefull applications
- Volume management
- HA applications with StatefulSets
How to secure your cluster
Production features
- CPU / RAM resource management
- Application self-scaling
- Manage your logs
- Helm
- The arrival of mesh services
Practical work
Conclusion and Key Take aways