CoreOS + Kubernetes Part by Part

This guide will use Tectonic Installer to deploy a Kubernetes cluster on Amazon AWS, then part it out, break the system, and watch it rebuild automatically.

These instructions require a CoreOS account, and an AWS account. See Downloading and installing Tectonic on AWS and Creating an AWS account for more information.

Cluster overview

Use Tectonic Installer to deploy Kubernetes clusters. Built on top of 100% open-source Kubernetes, Tectonic offers best practices solutions to cluster design and administration, including the following features:

• Highly Available Kubernetes: Easily deploy multiple masters and workers of Kubernetes. Governance and User Management: Govern with RBAC and integrate with your existing identity infrastructure: LDAP, SAML.
• Secure Networking: Set up secure networking policies with Flannel and Calico.
• Highly Available etcd with Disaster Recovery: Deploy HA etcd with built-in backup and restore capabilities.
• Automated Operations: Easily update and maintain your infrastructure with CoreOS’s Automated Operations.
• Container Linux Operating System: Leverage a lightweight Linux distribution, built for containers.
• Prometheus Monitoring and AlertManager: Monitor your cluster and applications with Prometheus. Create alerts using Tectonic Console.
• Logging & Auditing: Audit your infrastructure, tracking all API generated actions.

Deploy cluster

First, use Tectonic Installer to deploy a cluster on AWS using a GUI interface, or through the command line using Terraform.

Then, follow these guides to analyze, break, and watch your cluster rebuild.

• Set up kubectl
• Inspect the control plane
• Inspect a Kubernetes worker node
• Break system locally and watch recovery