Introduction

• Configuration management, service discovery, and distributed coordination
• etc vs traditional databases - documents (files) vs tables

How a Distributed System Works

• Load balancing, availability, recovery
• Nodes (master, slave, etc.)
• Messaging
• Storage

Overview of etcd Features and Architecture

• etcd in a cluster mode - leaders and followers
• Ensuring consensus using the Raft protocol

Setting up etc

• Downloading and running binaries
• Using the etcdctl CLI
• Configuring an etcd cluster

Operating etcd

• Connecting clients to the etcd service
• Adding and viewing "documents" using the etcd build-in client
• Reading and writing to etcd from within Kubernetes
• Reading and writing to etcd from a microservice application

Real-world Use Cases

• Changing application behavior in response to changes
• Using etcd to store database connection details
• Using etcd to store feature flags

Managing an etcd cluster

• Achieving fault tolerance
• Adjusting the number of nodes in cluster
• Reconfiguring node leader
• Performing distributed locking across a cluster
• Using metrics to track latency
• Backing up etcd data

etcd in the Public Cloud

• Cloud provider support (AWS, etc.)
• Networking pre-tuning

Best Practices

Troubleshooting

Summary and Conclusion