Skip to content

Abh10705/redis_rust

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

85 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

High-Performance Blockchain Cache (Powered by a Rust Core)

An intelligent, high-performance caching server designed to dramatically accelerate blockchain applications by reducing latency and offloading read requests from slow, expensive nodes.

This project implements a multi-threaded, Redis-compatible core in Rust to solve a real-world performance bottleneck for high-traffic dApps.

The Problem Querying data directly from a blockchain is inherently slow and costly. For high-traffic dApps, NFT marketplaces, and DeFi platforms, this latency creates a poor user experience and becomes a major scalability bottleneck.

This project solves that problem by providing a smart, in-memory caching layer that sits between the user and the blockchain, delivering instant responses for frequently accessed data.

Key Performance Indicator (KPI) Our primary metric for success is the Cache Hit Ratio, which measures the percentage of requests served directly from the cache.

Initial project success will be defined by achieving a >90% cache hit ratio under production load while securing an established blockchain firm as our first Lighthouse Customer to validate the solution.

Target Audience This project is designed for development teams building high-performance Web3 applications, including:

NFT Marketplaces (e.g., OpenSea, Magic Eden)

Decentralized Exchanges (DEXs) (e.g., Uniswap, Aave)

Blockchain Explorers (e.g., Etherscan, Solscan)

Web3 Gaming Platforms

Features Implemented Core Caching Engine:

String Operations: SET, GET, INCR with PX expiry for managing cache entries.

List Operations: LPUSH, RPUSH, LPOP, LLEN, LRANGE for caching ordered data like transaction histories or activity feeds.

Blocking Operations: BLPOP with timeout support for building real-time data pipelines.

System Architecture:

Concurrency: Fully multi-threaded, with each client connection handled in isolation. Shared state is managed safely with Arc<Mutex>.

Persistence: Loads database state from an .rdb file on startup to enable fast cache warming.

Transactions: Full MULTI, EXEC, DISCARD support for atomic operations.

High Availability & Scaling:

Replication: Can be deployed in a leader-follower configuration.

Performs the full replication handshake (PING, REPLCONF, PSYNC).

Master propagates all write commands to connected replicas to ensure cache consistency.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors