Date: May 19, 2016
By: Aaron Delp and Brian Gracely
Description: Brian talks with Stephen Fink (@sjfink, Distinguished Research Staff Member @IBM) about serverless computing, the architecture of IBM OpenWhisk, the open source community involvement and the use-cases he sees for this emerging technology.
- Get a free book from O'Reilly media or use promo code PCBW for a discount - 40% off Print Books and 50% off eBooks and videos
- Stephen Fink’s Background at IBM
- Stephen’s Research Page
- IBM OpenWhisk
- OpenWhisk Source Code on GitHub
- OpenWhisk on IBM BlueMix
- [Example - video processing] “Dark Vision”
- [Example - video processing] “Skylink”
- [Example - “Node Red” coordination for IoT devices]
- Want to register right now for VelocityConf? Use code PCCLOUD20 for 20% off registration.
Topic 1 - Welcome to the show. Give us some of your background in IBM’s Research group.
Topic 2 - “Events-based” and “Serverless” are terms that are starting to become more widely used. What types of challenges (use-cases) did you hear from developers that led you to create OpenWhisk?
Topic 3 - Let’s talk about the OpenWhisk architecture. The OpenWhisk model consists of three concepts:
- trigger: a class of events that can happen
- action: an event handler — some code that runs in response to an event
- rule: an association between a trigger and an action
Topic 3a - What are some of the really important things that people need to know about OpenWhisk? How it is different from other serverless frameworks that are starting to come into the market?
Topic 4 - Let’s walk through some of the early use-cases you’ve seen (or built).
Topic 5 - OpenWhisk has gotten some attention because it’s open-source. It’s still early, but where do you expect to see the most contribution - core OpenWhisk or at the edges around things that are specific to application factors?