10 个仓库
Infrastructure for managing large-scale distributed computing networks.
Distinguishing note: Focuses on the operational infrastructure layer.
Explore 10 awesome GitHub repositories matching devops & infrastructure · Distributed Infrastructure. Refine with filters or upvote what's useful.
Umami is a self-hosted, privacy-focused web analytics platform designed to provide full control over infrastructure and user data. It captures website traffic and visitor behavior through anonymous tracking methods that avoid cookies, browser fingerprinting, and the storage of personally identifiable information. The platform distinguishes itself through a comprehensive suite of behavioral analysis tools, including session replays, heatmaps, and cohort-based retention reporting. It features a multi-tenant architecture that allows teams to manage multiple websites within a single, collaborativ
Supports distributed architecture by running the web server and database on separate machines.
NATS Server is a high-performance, lightweight messaging system designed for cloud-native applications, edge computing, and distributed microservices. It functions as a distributed publish-subscribe broker that routes messages using hierarchical, dot-separated subject strings, enabling decoupled communication between services without requiring centralized broker lookups. The system supports core messaging patterns including asynchronous publish-subscribe, request-reply, and load-balanced queue processing. The platform distinguishes itself through a decentralized architecture that eliminates t
Connects separate messaging environments across cloud and edge systems to create a unified communication fabric.
TiKV is a cloud-native distributed transactional key-value store and storage engine. It provides a distributed database designed for horizontal scalability and strong consistency across a cluster of physical nodes. The system uses a Raft-based consensus mechanism to maintain data availability and state synchronization. It ensures ACID compliance for distributed transactions through a two-phase commit workflow and manages data distribution via multi-Raft sharding. The engine handles massive datasets using automated range splitting and cluster load balancing to distribute data across different
Processes data directly on storage nodes using a coprocessor to reduce network traffic and lower system latency.
PeerTube is a decentralized, open-source video hosting platform that enables users to operate independent, interoperable servers. By utilizing the ActivityPub protocol, it connects these servers into a global, federated network where users can follow channels, discover content, and interact across different instances. The platform is designed to function as a self-hosted video content management system, providing a community-driven alternative to centralized media services. What distinguishes PeerTube is its hybrid approach to content delivery and infrastructure management. It integrates peer
Delegates resource-intensive tasks to external infrastructure to maintain platform performance as content volume grows.
Owncast is a self-hosted live streaming server that provides full control over broadcast infrastructure and audience data. It functions as an RTMP video streaming server, accepting incoming video feeds and distributing them to viewers through HLS-based segmented streaming. The platform includes a built-in, stateful web-based chat interface that enables real-time viewer engagement during broadcasts. The project distinguishes itself through deep integration with the decentralized Fediverse, allowing servers to automatically broadcast stream status updates and notify followers across distributed
Bypasses internal re-encoding to reduce hardware utilization or offloads segment distribution to third-party storage providers.
ART is a platform for agentic training, providing a reinforcement learning framework, training environment, and compute orchestrator. It enables the improvement of multi-step agent reasoning and tool usage through group relative policy optimization and a judge-based reward modeling system. The project features tools for model distillation to transfer capabilities from large teacher models to smaller architectures, as well as a system for capturing execution trajectories to generate synthetic training data. It supports specialized training workflows including supervised fine-tuning for baselin
Offloads the training process to managed hardware to remove the requirement for local GPU resources.
Performs operations like AllReduce directly on network switches to reduce CPU involvement and speed up multi-GPU training.
Orchestrates multiple Xray-core servers as nodes to distribute proxy traffic for redundancy and scalability.
caddy-docker-proxy 是一个动态 HTTP 反向代理和 Docker 网络入口控制器,通过读取 Docker 容器的标签自动生成路由配置。它作为服务发现工具,实时检测容器 IP 地址,将传入的 Web 流量路由到正确的后端目标。 该项目是一个分布式代理编排器,能够将生成的配置从中央控制器推送到多个远程服务器实例,以扩展请求处理能力。它自动化了代理域名的 TLS 安全证书颁发和续期,并协调服务器副本之间的共享证书。 该系统支持基于标签的配置生成,使用模板和 Docker secrets 将元数据和全局设置注入服务器配置。它通过监听容器生命周期事件来保持同步,并利用进程内加载机制更新服务器行为。 流量管理功能包括定义请求路由规则、路径匹配和 URI 重写,以将流量引导至容器化资源。
Coordinates multiple proxy server instances from a central controller to scale request handling across a cluster.
该项目是一个可扩展的实时通信平台,专为多方视频和音频会议而设计。它作为一个模块化媒体服务器,编排实时会话、管理虚拟会议室,并处理跨分布式计算环境的媒体流分发。 该平台通过支持协议桥接脱颖而出,允许在 WebRTC、RTSP、SIP 和 HLS 之间进行转换。它结合了硬件加速转码以在不同设备上保持性能,并利用基于插件的架构来实现自定义媒体处理,例如实时内容分析和对象检测。 该系统为管理媒体管道提供了全面的基础设施,包括会话录制、动态网络适配以及自定义视频布局的组合。它还支持电话集成,实现数字会议室与传统电话系统之间的连接。管理功能包括多租户服务管理和基于令牌的访问控制,以保护虚拟空间。 该软件已打包用于容器化部署,促进了跨开发和生产集群的一致执行和编排。
Manages virtual rooms and horizontal scaling of media workloads across distributed computing clusters.