Skip to content

etcd.md

Latest commit

 

History

History
47 lines (30 loc) · 1.75 KB

etcd.md

File metadata and controls

47 lines (30 loc) · 1.75 KB
Raw

🗃️ etcd – The Backend Database of Kubernetes

🔹 What is etcd?

etcd stands for "/etc distributed", inspired by the Unix /etc directory for configurations — etcd is a distributed key-value store used as the backing store for all Kubernetes cluster data.

It is the source of truth for the entire Kubernetes control plane. Every object you create, modify, or delete in Kubernetes — such as Pods, ConfigMaps, Services, etc. — is stored in etcd.

🔹 Core Characteristics

Feature Description
Distributed Can run as a cluster of nodes
Consistent Uses the Raft consensus algorithm
Highly Available Supports HA setups with odd-numbered members
Strongly Consistent Ensures linearizable reads and writes
Secure (TLS-enabled) Communicates over encrypted channels

🔹 Responsibilities of etcd in Kubernetes

  1. Persistent Storage

    • Stores the entire cluster state — including all objects and their current status.

  2. High Availability and Fault Tolerance

    • Can tolerate failure of minority members in an HA setup.

  3. Data Consistency

    • Guarantees strong consistency using the Raft consensus algorithm.

  4. Watch and Notification System

    • Allows components (like kube-apiserver) to watch keys and be notified when data changes.

  5. Key-Value Store Only

    • Doesn't understand Kubernetes natively; it's just a secure and reliable data store.

🔹 How it Works with Kubernetes

  • kube-apiserver is the only component that directly interacts with etcd.
  • Controllers, schedulers, and other components watch the API server — not etcd.