Home

🌌 Welcome to the VortexKV Wiki

VortexKV (Vector-Optimized Redis-Compatible Throughput Extreme Key-Value Store) is a next-generation, cyberpunk in-memory data store engineered from scratch in pure Go. It delivers 6,870,000+ ops/sec pipelined throughput (with peak bursts up to 9,411,764 ops/sec) and 210,000+ ops/sec direct concurrency with ~111µs p50 latency, hardware-accelerated kqueue/epoll event reactor, 64 cacheline-padded lock-striped shards, smart socket write coalescing, native HNSW AI vector graphs, Redis Streams with consumer groups, an autonomous cluster gossip bus, dual scripting (Lua 5.1 & Wasm), and an embedded visual command deck.

---

🧭 Wiki Table of Contents

Topic	Description	Link
🚀 Getting Started	Installation via cURL, Docker, Helm, or source	Getting-Started
🏎️ Architecture & Core	64 lock-striped shards, memory models, zero-alloc timing wheel	Architecture-and-Internals
🧠 Native AI Vector Search	HNSW skip-graphs, Cosine/Euclidean/Dot similarity (VSEARCH)	AI-Vector-Search
🌊 Streams & Consumer Groups	Event streaming, Pending Entries List (PEL), and worker pools	Streams-and-Consumer-Groups
📡 Cluster & Gossip Protocol	16,384 hash slots, Port 17379 binary bus, autonomous failover	Cluster-and-Gossip-Protocol
⚖️ Cluster Slot Rebalancer	CLI automation for live online slot and key migrations	Cluster-Rebalancing
🔁 Replication & Failover	Master-Replica async replication, PSYNC handshake, instant failover	Replication-and-Failover
📜 Scripting Engines	Embedded Lua 5.1 & WebAssembly (Wasm) runtime via Wazero	Scripting-Lua-and-Wasm
☸️ Kubernetes Operator	Declarative CRD (kind: VortexCluster) and auto-scaling	Kubernetes-Operator
💾 Persistence & Durability	Binary RDB snapshots with CRC64 & durable Append-Only File (AOF)	Persistence-and-Durability
🔒 Security & ACL	Redis 6+ multi-user access control, TLS encryption, and tokens	Security-and-ACL
📊 Observability	Prometheus /metrics exporter and /healthz Kubernetes probes	Telemetry-and-Metrics

---

⚡ Core Philosophy & Architecture

                       ┌───────────────────────────────┐
                       │   Client Application / SDK    │
                       │ (Python, Node, Go, redis-cli) │
                       └───────────────┬───────────────┘
                                       │ RESP2 / RESP3 (Port 7379)
                                       ▼
                       ┌───────────────────────────────┐
                       │   Vortex Wire Parser & Auth   │
                       └───────────────┬───────────────┘
                                       │
            ┌──────────────────────────┴──────────────────────────┐
            ▼                                                     ▼
┌───────────────────────┐                             ┌───────────────────────┐
│ 64-Shard Lock Striped │                             │ Embedded Visual Deck  │
│  Concurrent Keyspace  │                             │   (HTTP Port 7380)    │
└───────────┬───────────┘                             └───────────────────────┘
            │
            ├──────────────► 🧠 HNSW Vector Index (Multi-layer Skip Graphs)
            ├──────────────► 🌊 Event Streams Engine (PEL & Consumer Groups)
            ├──────────────► 📜 Lua 5.1 & Wasm Engine (Pure-Go Sandboxed)
            ├──────────────► 📡 Gossip Bus & Cluster Mesh (Port 17379)
            └──────────────► 💾 Dual Durability (AOF + Binary RDB Snapshots)

1. Breaking the Global Mutex Bottleneck

Standard Redis processes all commands through a single thread to avoid concurrency issues, limiting throughput to a single CPU core. VortexKV pairs a hardware-accelerated Multi-Reactor engine (kqueue/epoll) with 64 independent cacheline-padded lock-striped shards and smart socket write coalescing, allowing high-concurrency workloads to utilize all CPU cores simultaneously and batch pipelined responses into consolidated kernel writes (6.8M+ ops/sec peak, 2.7M+ ops/sec GET).

2. Dedicated Non-Conflicting Ports

VortexKV is engineered for seamless coexistence with existing database infrastructure:

• 7379: RESP2/RESP3 Redis client wire protocol (leaves standard port 6379 open).
• 7380: Embedded Cyberpunk Web Studio & HTTP REST API.
• 17379: Cluster binary heartbeat and gossip bus (client_port + 10000).

---

⚡ 5-Second Quick Verification

# Connect via standard redis-cli:
redis-cli -p 7379 -a "vortex_secure_2026" PING
# PONG

# Store high-dimensional embeddings:
redis-cli -p 7379 -a "vortex_secure_2026" VADD embeddings doc_ai 0.95 0.05 0.0 0.0

# Search Top-1 nearest neighbor using Cosine similarity:
redis-cli -p 7379 -a "vortex_secure_2026" VSEARCH embeddings 1 cosine 0.90 0.10 0.0 0.0
# 1) 1) "doc_ai"
#    2) "0.999512"

---

🔗 External Links

• GitHub Repository: GargAnshu9468/vortexkv
• Live Web Sandbox: garganshu9468.github.io/vortexkv
• License: MIT Open Source