Discovering distributed task queue in its simplest form using Retask
by Kushal Das (speaking)
Objective
The primary objective of this talk is to present Retask as a distributed Task/Job queues system.
Retask uses Redis in the backend and keeps the data in JSON format.
After the talk people will have the knowledge required
to start using Retask in their own projects as distributed queue.
Description
Ever thought of having a distributed task queue ready to be used and yet very simple ?
Retask is a distributed task queue system written in Python. It has one of the simplest API ever
written for a distributed task queue. It supports both Python 2.x and 3.x.
The talk will demonstrate:
- Using queues can be easy!
- API (5 tiny methods, I promise)
- Asynchronous & synchronous communication between workers and producers.
- Real life use cases in Fedora Infrastructure and inside Red Hat.
Talk outline
- Introduction to Retask
- Retask is not a message queue. It is not AMQP or zmq.
-
Retask is for creating Task/Job queues (All above is for the basic introduction of the library and reason I started writing it). ^^ 8 minutes
-
Having that simplest queue (First example demoed.)
-
Asynchronous result passing example, showing non-blocking calls from Retask. (We will have 1 producer and two workers, workers will return some result data in async non blocking calls) ^^ 5 minutes
-
Synchronous result passing example, showing blocking calls from Retask. (Using blocking calls, we will try the same example shown above.) ^^ 5 minutes
-
Use cases inside Fedora Infrastructure: Async Caching in Fedora Packages App & Darkserver project.
- Use cases inside Red Hat
^ ^ 5 minutes
- Performance tips
^^ 2 minutes
Rest for QA
Speaker bio
Kushal Das is a core developer of CPython. He is a Fedora Ambassador and also director at Python Software Foundation.
He is a regular speaker in different technical conferences including previous PyCon(s) and also does Python workshops in various events & engineering colleges.
He also wrote a beginner Python book called Python for you and me which is targeted to the students for fast paced learning and available freely online.
1
▼
Link to retask in description is broken. It is pointing to same proposal.