Skip to main content

Simple, explicit durable execution framework on top of ZeroMQ and SQLModel.

Project description

Pipeteer

Simple, explicit durable execution framework.

Read the docs

Welcome to Pipeteer

Pipeteer simplifies the complexity of durable execution whilst not hiding the underlying persistence.

Why Pipeteer?

Use pipeteer if you need...

  • Persistance: your app can stop or crash and resume at any time without losing progress
  • Observability: you can see the state of your app at any time, and modify it programmatically at runtime
  • Exactly-once semantics: your app can be stopped and resumed without dropping or duplicating work
  • Fault tolerance: if a task fails, it'll keep working on other tasks and retry it later
  • Explicitness: pipeteer's high level API is a very thin abstraction over SQLModel (for persistance) and ZeroMQ (for inter-process communication)

Proof of Concept

Definition. You can define a durable workflow this easy:

from pipeteer import activity, workflow, WorkflowContext

@activity()
async def double(x: int) -> int:
  return 2*x

@workflow()
async def quad(x: int, ctx: WorkflowContext) -> int:
  x2 = await ctx.call(double, x)
  x4 = await ctx.call(double, x2)
  return x4

Worker. And here's how to run it:

import asyncio
from pipeteer import DB, Context

db = DB.at('pipeline.db')
ctx = Context.of(db)

async def main():
  await asyncio.gather(
    double.run(ctx),
    quad.run(ctx),
    ctx.zmq.proxy(),
  )

Input. How to give it tasks?

from pipeteer import DB, Context

db = DB.at('pipeline.db')
ctx = Context.of(db)

Input = quad.input(ctx)
with db.session as s:
  s.add(Input(key='task', value=1, output='my-output'))
  s.commit()

await quad.notify(ctx)

Output. How to get the results?

from sqlmodel import select
from pipeteer import DB, Context

db = DB.at('pipeline.db')
ctx = Context.of(db)

Output = quad.output(ctx, 'my-output')
while True:
  with db.session as s:
    for entry in s.exec(select(Output)):
      print(f'Output: {entry.key} -> {entry.value}')
      s.delete(entry)
    s.commit()
  await ctx.wait('my-output')

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pipeteer-1.0.2.tar.gz (11.7 kB view details)

Uploaded Source

Built Distribution

pipeteer-1.0.2-py3-none-any.whl (15.7 kB view details)

Uploaded Python 3

File details

Details for the file pipeteer-1.0.2.tar.gz.

File metadata

  • Download URL: pipeteer-1.0.2.tar.gz
  • Upload date:
  • Size: 11.7 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.1.1 CPython/3.11.10

File hashes

Hashes for pipeteer-1.0.2.tar.gz
Algorithm Hash digest
SHA256 9a3fa2de735507f289dbc50181595c87b67e1d91b4d630ae3468bb6d9c85f64d
MD5 601150d3cc3d3273df9406b306a4cbe5
BLAKE2b-256 ec78eff36940bcef4aad64ee677b22938022e299d4107461183b7a2a54225572

See more details on using hashes here.

File details

Details for the file pipeteer-1.0.2-py3-none-any.whl.

File metadata

  • Download URL: pipeteer-1.0.2-py3-none-any.whl
  • Upload date:
  • Size: 15.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.1.1 CPython/3.11.10

File hashes

Hashes for pipeteer-1.0.2-py3-none-any.whl
Algorithm Hash digest
SHA256 4f4656bef27115c4360d4143e25fe4ba86cc4ce14319f3d6a7ad73f3235ce486
MD5 0d33b96b4a9261fff4245abc4eef6288
BLAKE2b-256 dbde32724e86c398dfcea616efd9214f002becf38356bd80fed3ddc4ea30adc0

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page