[KongchangAI]
Product

etcd

An open-source distributed key-value store based on the Raft consensus algorithm, commonly used for service discovery, configuration management, and distributed coordination.

Core Facts

Timeline (last 90 days)

Nov 7

Milvus integrates ETCD for metadata management and service discovery, uses message middleware such as Kafka or Pulsar to handle log streams, and its object storage supports Amazon object storage, MinIO and various cloud storages.

Verified65%
Nov 6

etcd is a CNCF graduated project with a mature community and broad ecosystem integration.

Unverified50%
Nov 6

etcd's transactions support version-number-based CAS (compare-and-swap) semantics, which is the key to implementing optimistic concurrency control.

Unverified50%
Nov 6

Distributed locks are usually implemented using etcd's transactions (Txn) and leases: a client checks in a transaction whether a key does not exist (by comparing the version number to 0), and if so writes and binds a lease, thereby atomically acquiring the lock; the lock holder must keep renewing, and after a crash the lease expires and the lock is automatically released.

Unverified50%
Nov 6

Kubernetes uses etcd to store the state of all cluster objects, which is etcd's most well-known application.

Unverified50%
Nov 6

etcd supports TLS-encrypted communication and role-based access control (RBAC), and also provides comprehensive monitoring metrics and backup and recovery tools.

Unverified50%
Nov 6

watch is implemented based on etcd's MVCC (multi-version concurrency control): each modification generates a new revision number, and clients can specify a revision to start watching from, thereby supporting replay of historical events and no event loss after reconnection.

Unverified50%
Nov 6

A lease is a TTL-based authorization mechanism in etcd: a client applies for a lease and binds several keys to it, and must renew it periodically (keepalive); once the lease expires or the client loses contact, the bound keys are automatically deleted.

Unverified50%
Nov 6

etcd provides HTTP/gRPC interfaces and supports key-value create, read, update, delete, watch, and lease mechanisms.

Unverified50%
Nov 6

Raft guarantees strong consistency (linearizable reads and writes), while high availability depends on a majority of nodes being alive: when the cluster loses a majority of nodes, etcd rejects writes to protect consistency, which is a typical CP choice in the CAP trade-off.

Unverified50%

22 more timeline events

All Facts (20)

Verified

Milvus integrates ETCD for metadata management and service discovery, uses message middleware such as Kafka or Pulsar to handle log streams, and its object storage supports Amazon object storage, MinIO and various cloud storages.

65%
Unverified

Every write operation to etcd goes through Raft log replication where the Leader receives the request, replicates to a majority of Followers, and returns success after confirming the commit

90%
Unverified

Milvus架构中etcd承担分布式协调角色,存储集群元数据和节点状态

85%
Partially Verified

etcd is the core data store for Kubernetes and uses Raft to guarantee strong consistency of cluster configuration data

65%
Partially Verified

Docker、Kubernetes、Terraform、etcd、Prometheus等云原生核心项目几乎都用Go编写

65%
Unverified

etcd is a CNCF graduated project with a mature community and broad ecosystem integration.

50%
Unverified

etcd implements strong consistency based on the Raft consensus algorithm, ensuring that the cluster can still read and write normally as long as a majority of nodes are alive.

50%
Unverified

etcd recommends deploying an odd number of nodes (usually 3 or 5), because writes require confirmation from more than half of the nodes.

50%
Unverified

Raft and Paxos have similar goals but Raft is easier to implement in engineering, and etcd's raft library is reused by many projects.

50%
Unverified

Raft guarantees strong consistency (linearizable reads and writes), while high availability depends on a majority of nodes being alive: when the cluster loses a majority of nodes, etcd rejects writes to protect consistency, which is a typical CP choice in the CAP trade-off.

50%
Unverified

etcd provides HTTP/gRPC interfaces and supports key-value create, read, update, delete, watch, and lease mechanisms.

50%
Unverified

A lease is a TTL-based authorization mechanism in etcd: a client applies for a lease and binds several keys to it, and must renew it periodically (keepalive); once the lease expires or the client loses contact, the bound keys are automatically deleted.

50%
Unverified

watch is implemented based on etcd's MVCC (multi-version concurrency control): each modification generates a new revision number, and clients can specify a revision to start watching from, thereby supporting replay of historical events and no event loss after reconnection.

50%
Unverified

etcd supports TLS-encrypted communication and role-based access control (RBAC), and also provides comprehensive monitoring metrics and backup and recovery tools.

50%
Unverified

Kubernetes uses etcd to store the state of all cluster objects, which is etcd's most well-known application.

50%
Unverified

Distributed locks are usually implemented using etcd's transactions (Txn) and leases: a client checks in a transaction whether a key does not exist (by comparing the version number to 0), and if so writes and binds a lease, thereby atomically acquiring the lock; the lock holder must keep renewing, and after a crash the lease expires and the lock is automatically released.

50%
Unverified

etcd's transactions support version-number-based CAS (compare-and-swap) semantics, which is the key to implementing optimistic concurrency control.

50%
Unverified

In systems such as Kubernetes, when etcd or the API Server is unavailable, Pods may still continue running on nodes, but users can no longer issue new instructions.

50%
Unverified

Systems such as etcd, Consul, and TiKV all use the Raft consensus algorithm.

50%
Unverified

In distributed systems, equivalents of CAS include Redis's SET NX and etcd's compare-and-swap transaction.

50%

Source Articles