Last summer, 3 new parallel and distributed computing projects for Python were open-sourced:

• Scaler, a light-weight and resilient distributed task scheduler for Python;
• Parfun, an easy to use map-reduce decorator for Python;
• Pargraph a battle-tested distributed graph engine for Python.

These packages have been used in production by a large US financial institution.

In this short talk, Raphael will share his experience taking part to the development of these 3 projects, and how to leverage them for highly distributed workloads.