lni/dragonboat
Dragonboat
Dragonboat هو تنفيذ بلغة Go لبروتوكول إجماع Raft مصمم للحفاظ على حالة متسقة عبر مجموعة موزعة من العقد. يوفر مكتبة لبناء آلات حالة موزعة تضمن سلامة البيانات وتحمل الأخطاء أثناء فشل النظام.
يتميز المشروع بتنفيذ Raft متعدد المجموعات، والذي يقوم بتقسيم البيانات عبر مجموعات إجماع مستقلة لتوزيع أعباء العمل وزيادة سعة معالجة النظام الإجمالية. كما يدمج بروتوكول TLS المتبادل لتشفير الاتصال بين العقد والتحقق من هوية أعضاء المجموعة.
يتضمن النظام قدرات لآلات حالة عالية الأداء مع دعم لكل من الذاكرة وتخزين القرص. يتميز بتحسينات مسار القراءة لضمان الاتساق دون توليد إدخالات سجل جديدة، وواجهة تخزين قابلة للتوصيل لخلفيات السجل المخصصة، وأدوات إدارية لاستعادة النصاب القانوني لاستعادة التوفر بعد فقدان دائم لأغلبية العقد.
يتم دعم الاستقرار التشغيلي من خلال تصدير مقاييس صحة المجموعة.
Features
- Raft Consensus Implementations - Provides a Go implementation of the Raft protocol to manage replicated logs and leader election for distributed consistency.
- Multi-Raft Sharding - Utilizes a multi-Raft sharding architecture where independent consensus groups manage different data ranges to scale throughput.
- Distributed Consensus Algorithms - Implements protocols that ensure agreement on a sequence of operations across network nodes to maintain consistent state.
- Distributed Coordination Systems - Provides a framework for managing state, synchronization, and consensus across distributed nodes to ensure data consistency.
- Distributed State Machines - Provides a library for building distributed state machines that maintain consistency and data integrity across a cluster.
- Distributed State Synchronization - Provides a mechanism for synchronizing application state across distributed server instances to ensure fault tolerance.
- Replicated State Machines - Implements replicated state machines that maintain identical state across nodes using an ordered sequence of log operations.
- Quorum Restorations - Includes a utility for restoring quorum and designating a new leader after a cluster has lost its majority of voting members.
- Consistent Read Optimizations - Implements read-path optimizations that verify the latest committed index to ensure consistency without generating new log entries.
- Durable State Machine Persistence - Supports building durable application states that persist to disk or reside in memory for fast recovery.
- Pluggable Storage Backends - Provides a pluggable storage interface allowing the consensus log to be persisted in various database backends.
- Quorum Recovery Mechanisms - Provides administrative tooling to restore cluster availability by manually reconfiguring membership after a permanent majority loss.
- Consistent Read Protocols - Implements a specialized index protocol for read-only queries to ensure strong consistency without appending new entries to the transaction log.
- Consensus Log Storage Integrations - Supports integrating custom storage backends and transport layers to optimize performance for the consensus log.
- Cluster Communication Security - Enforces mutual TLS authentication across internal cluster components to ensure secure communication in open environments.
- Mutual TLS Authentication - Secures inter-node communication using mutual TLS to encrypt traffic and verify member identity.
- Mutual TLS Transports - Secures inter-node communication using mutual TLS transports for encrypted tunnels and certificate-based identity verification.
- State Persistence Layers - Supports both memory and disk persistence for application state to ensure durability across system restarts.
- Consensus Cluster Recoveries - Offers mechanisms to restore availability to a consensus cluster after a majority of nodes are permanently lost.
- Distributed Systems - High-performance library for distributed consensus.
- Algorithms - Raft consensus algorithm library.
- Distributed Systems - High-performance multi-group Raft consensus library.
