Language: English (with English Slides)
Developer and starting software architect in cloud-native application development, founder of Project Invader open source organization, mainly focused on developing and designing platform and software as a service applications for OpenStack. He’s lead software developer in EPAM, concentrating on product development along with bringing well-designed and production ready integration with clouds environments, including vCloud Air and vSphere. He is a contributor to OpenStack DBaaS platform and OpenStack CloudValidation open source framework. Founder and technical project lead in AIOrchestra, BeeDB project.
AIOrchestra is the way to rethink the way we develop and deploy. The idea that stands behind it - open platform, open capabilities, open TOSCA. While AIOrchestra being an engine to process TOSCA models into executable deployment graph using Python 3.5 and its asynchronous API. The hardest part of AIOrchestra was to convert TOSCA models into an executable object. Based on TOSCA understanding it was required to have asynchronous tasks that may run in parallel but having common synchronization point between them. Starting version 3.4, Python standard library provides asynchronous API using event-driven handling concept through coroutines. This is exact technology above which AIOrchestra is built. So, this talk will discover exact lesson that we’ve learnt while building dynamic graph of sequenced asynchronous tasks to accomplish concrete result - do a consistent asynchronous deployment orchestration.
Tags: TOSCA orchestration, AsyncIO, OpenStack