hashicorp/raft

Features

• Consensus Algorithms - Implements the Raft consensus algorithm to ensure a majority of nodes agree on a single system state.
• Distributed State Synchronization - Synchronizes replicated logs across a cluster of servers to ensure all nodes agree on the system state.
• Raft Consensus Implementations - Provides a complete Go implementation of the Raft consensus protocol for managing replicated logs and leadership.
• Consensus Engines - Acts as the core engine managing state replication and consensus for a group of servers.
• Distributed Commit Logs - Monitors replicated log indexes and notifies when a majority of servers have persisted an entry.
• Commit Index Tracking - Determines the highest log entry replicated on a quorum of servers to ensure consistent state commitment.
• Single-Leader Replication - Centralizes log management by designating a single leader node to coordinate all client write operations.
• Bespoke Distributed System Primitives - Provides low-level coordination primitives in Go for building custom fault-tolerant distributed systems.
• Distributed Leader Election - Coordinates a distributed leader election process to ensure a single active instance manages the replicated log.
• Fault Tolerance - Designs distributed services that maintain data consistency and availability during node failures or network partitions.
• Consensus Log Replication - Sends log entries to a majority of peers and applies them to the state machine for consistency.
• Replicated Log Management - Synchronizes a log of operations across servers to ensure all nodes agree on the sequence of changes.
• Consensus Node Roles - Tracks and updates the operational role of nodes using thread-safe operations to maintain cluster status.
• Server State Synchronization - Synchronizes data across a cluster of servers to ensure all nodes converge on a shared log and state.
• Consensus and Coordination Systems - Implements core consensus algorithms to maintain a consistent global state across distributed nodes.
• Consensus Protocols - Routes protocol requests over TCP to ensure state synchronization and propagate changes from leader to followers.
• Node Failure Detection - Implements failure detection using periodic heartbeats to trigger new leader elections when the current leader fails.
• Inter-Node Consensus Communication - Facilitates the exchange of voting and log replication requests between peer servers to maintain cluster consensus.
• Resource State Synchronization - Synchronizes replicated logs across multiple servers using a consensus protocol to ensure global state agreement.
• Replicated State Machines - Provides a replicated state machine that applies an ordered sequence of operations deterministically across a cluster.
• Configuration Persistence - Persists key configuration values to stable storage to ensure system safety across restarts.
• Point-In-Time Snapshots - Writes point-in-time copies of the state and configuration to disk for faster recovery and log truncation.
• Local Consensus Snapshots - Writes system state snapshots to the local filesystem and manages retained versions to prevent disk exhaustion.
• Buffer Caches - Stores specific log entries in a fast-access ring buffer to reduce read latency from the primary store.
• In-Memory Caches - Uses an in-memory ring buffer to cache recently written log entries and reduce disk I/O.
• Log Truncation - Discards snapshot data during the snapshotting process to truncate the replicated log.
• Pluggable Storage Drivers - Offers a pluggable architecture to decouple consensus logic from physical storage implementations for logs and snapshots.
• Snapshot Synchronization - Implements fast synchronization of nodes using state snapshots to bootstrap the cluster state.
• State Checkpointing - Triggers automatic state snapshots based on time intervals or log size thresholds to enable reliable recovery.
• Log State Recovery - Provides mechanisms to restore a node's state by replaying committed log entries or loading snapshots.
• Log Compaction - Includes log compaction to reduce storage by persisting the current state and discarding superseded log entries.
• Cluster Management - Tracks a list of servers and their voting roles using unique identifiers to prevent configuration conflicts.
• Leadership Transfers - Orchestrates the movement of the leader role to another peer to facilitate maintenance or controlled failover.
• Failure Simulation Tools - Injects network partitions and transport errors to verify distributed system recovery from node isolation.
• Heartbeat Monitors - Implements a dedicated heartbeat handler to maintain cluster liveness without interference from disk latency.
• Transport Simulation - Creates an in-memory transport layer to test consensus logic and message passing without network connectivity.
• Node Bootstrapping - Transfers state snapshots to peer nodes to quickly synchronize their logs and state machines.
• Cluster State Captures - Provides point-in-time snapshots of the state machine to truncate logs and accelerate recovery for lagging nodes.
• Cluster Bootstrapping Mechanisms - Manages cluster bootstrapping by defining participating peer addresses and initial member configurations.
• Application Batching - Processes multiple state machine applications in groups to increase overall throughput and performance.
• Cluster Health Monitoring - Provides interfaces to query leader addresses, peer states, and synchronization indices to monitor cluster health.
• Cluster Monitoring Systems - Streams real-time cluster events, such as leadership changes and heartbeat failures, to monitoring observers.
• Consensus Verification - Validates that replicated logs and state machine sequences remain consistent across all nodes after operations.
• Deterministic State Testing - Uses mock state machines and in-memory configurations to verify log replication and state consistency.
• Distributed Systems - Implementation of the Raft consensus protocol.
• Distributed Systems and RPC - Implementation of the Raft consensus protocol for distributed systems.
• Algorithms - Raft algorithm implementation.
• Distributed Systems - Raft consensus protocol implementation.

Star history