◉ COOLJAPAN
← All posts
March 28, 2026 • 4 min read

CeleRS 0.2.0 Released — Pure Rust Celery-Compatible Distributed Task Queue

Production-ready, binary-level compatible drop-in replacement for Python Celery. 18 crates (100% complete), 4,075 tests, 10× throughput, type-safe macros, 5 brokers + 3 backends, Canvas workflows, DLQ, priority queues, task cancellation, and full observability. <50 MB memory, 10,000 tasks/sec per worker. The sovereign distributed task queue layer for SciRS2 and the entire COOLJAPAN ecosystem (now 26M+ SLoC total).

release celers celery task-queue distributed-systems pure-rust workflow scirs2 enterprise

The distributed task queue foundation of the COOLJAPAN ecosystem just reached full production readiness.

Today we released CeleRS 0.2.0 — a complete, production-grade pure Rust distributed task queue library that delivers binary-level protocol compatibility with Python Celery while being dramatically faster, type-safe, and enterprise-ready.

No Python. No GIL. No fragile worker processes.
No C/C++ dependencies. No runtime surprises.
Just clean, memory-safe, high-throughput task execution that compiles to a single static binary and runs everywhere — from laptops to edge servers to massive cloud clusters.

Why CeleRS 0.2.0 is a game changer

For years, reliable distributed task queues in enterprise environments meant depending on Python Celery — powerful but plagued by interpreter overhead, memory leaks, and limited type safety.

These tools are powerful but suffer from:

CeleRS 0.2.0 ends all of that.

It delivers 10× throughput compared to Python Celery while being 100% memory-safe and fully type-checked at compile time.
Notable results:

Technical Deep Dive: How We Built a Production-Grade Celery Replacement in Pure Rust

CeleRS follows a clean, layered architecture inspired by Celery but rebuilt from the ground up for modern Rust:

  1. Core & Protocol Layer (celers, celers-core, celers-protocol, celers-kombu)
    Full binary-compatible Celery v2/v5 wire protocol + procedural macros (#[celers::task]) for compile-time verified task signatures.

  2. Broker Support (5 high-performance backends)
    Redis (Lua + pipelining), PostgreSQL (ACID-safe FOR UPDATE SKIP LOCKED), MySQL (batch ops), RabbitMQ (AMQP exchanges), AWS SQS (long polling).

  3. Result Backends (3 options)
    Redis (TTL + chord sync), Database (PostgreSQL/MySQL with analytics), gRPC (microservices-native).

  4. Workflow & Scheduling (celers-canvas, celers-beat)
    Full Canvas primitives: Chain, Group, Chord, Map, Starmap, Signature. Periodic tasks via cron, interval, and solar triggers.

  5. Enterprise Features
    Priority queues, Dead Letter Queue (DLQ) with inspection/replay, task cancellation via Pub/Sub, exponential backoff retries, timeout enforcement, graceful shutdown, and event persistence.

Key Rust advantages:

What’s inside 0.2.0 (released March 28)

This is the foundation

CeleRS is now the official distributed task queue backend for the entire COOLJAPAN stack (total ecosystem: 26M+ SLoC Rust, 660+ crates, 40+ production-grade libraries):

Repository: https://github.com/cool-japan/celers

Star the repo if you want a type-safe, high-performance task queue without Python’s overhead.

The era of “just run Celery workers in Python” is over.

Pure Rust distributed task queues — binary compatible, 10× faster, and fully sovereign — are here.

— KitaSan at COOLJAPAN OÜ March 28, 2026

↑ Back to all posts