January 28, 2026
Task queues, hot takes
Oban, the job processing framework from Elixir, has come to Python
Python gets Oban: cheers, side‑eye at Celery, and a Sidekiq cameo
TLDR: Oban, a popular Elixir tool for running background tasks, has landed in Python with a free tier and a paid Pro upgrade. The crowd is buzzing over database‑first simplicity, a Sidekiq creator cameo, Celery fatigue, and open‑core controversy—making this a big deal for teams choosing their next job runner.
Python just got Elixir’s cult‑favorite job runner, Oban—and the crowd went instantly split. Celery veterans (that’s the popular Python task tool) shuffled in with “it works, but ouch,” while one dev joked they’re ready to swap their celery for something less wilted. Another camp shouted that this database‑first approach is the glow‑up they’ve been begging for: keep your tasks in Postgres, not a separate cache, and stop babysitting extra infrastructure.
Then Sidekiq’s own creator, mperham, dropped in like a surprise guest star, noting Oban’s roots in Sidekiq and plugging his cross‑language alternative, Faktory. Cue the framework family drama. Meanwhile, workflow‑tool skeptics said Temporal is “too much hammer for our nails,” making Oban’s “lighter” vibe sound very attractive.
But the real spice? Open‑core angst. Oban’s free version is fine for small apps, while Pro unlocks the fast lane—true parallelism, bulk operations, smarter recoveries. Some cheered the sustainability, others side‑eyed the paywall. One commenter summed it up: “I support it in theory, but I hesitate in practice.” Over in the peanut gallery, Elixir diehards begged for the reverse migration: “Why not bring Python’s data pipelines to Elixir?” Translation: if Elixir’s the concurrency king, why is Python getting the party? Between database‑vs‑Redis debates, “Celery detox” jokes, and a Sidekiq shout‑out, Oban’s Python debut is less tech launch, more reality‑TV reunion—with code.
Key Points
- •Oban’s Python implementation (Oban-py) brings a database-centric job processing model to Python.
- •Oban-py offers two editions: open source with limitations and a Pro version that lifts them and adds features.
- •OSS Oban-py uses single-threaded asyncio, lacks bulk inserts/acks, and may inaccurately rescue long-running jobs.
- •Pro features include smarter heartbeats, bulk operations, workflows, relay, unique jobs, and smart concurrency.
- •Under the hood, jobs are inserted into PostgreSQL (oban_jobs), NOTIFY awakens stagers, producers receive events, and dispatch jobs to workers.