---
name: "dragonflydb-multithreaded-cache"
description: "Achieve millions of QPS per node using modern multi-threaded share-nothing DragonflyDB."
metadata:
  author: "DragonflyDB"
  version: "1.0.0"
  runtime_compatibility:
    - "codex"
    - "claude-code"
    - "cursor"
    - "windsurf"
triggers:
  - "deploy dragonflydb"
  - "benchmark dragonfly vs redis"
  - "configure dragonfly cache tiering"
permissions:
  fs_read: true
  fs_write: "workspace"
  shell_exec: false
  net_access: false
---

# DragonflyDB Multi-Threaded Redis Alternative

Optimizes high-volume memory workloads with DragonflyDB: utilizing modern hardware architecture, vcache memory tiering, zero fork latency snapshots, and 25x higher throughput than legacy Redis instances.

## When to Invoke
Invoke this skill whenever you need to:
- deploy dragonflydb
- benchmark dragonfly vs redis
- configure dragonfly cache tiering

## Execution SOP & Rules
1. Run DragonflyDB: docker run -p 6379:6379 --ulimit memlock=-1 docker.dragonflydb.io/dragonflydb/dragonfly.
2. Configure memory flags: --maxmemory=32GB --cache_mode=true for pure LRU eviction.
3. Connect using standard redis-cli or application ORM clients without modifying client code.
4. Monitor per-thread performance and memory efficiency via INFO ALL command output.

## Verification Checklist
- Run local linting, syntax validation, and type-checks before reporting completion.
- Ensure all created or modified files adhere to workspace boundaries and avoid uncommitted secrets.
- Verify tests or build commands exit cleanly with return code 0.