Why?
1. No vendor locking
2. Any language. Even a command
line tool can be exposed
3. not limited to 300 seconds as
with Lambda
4. Run it wherever you want(rpi)
5. Get used to Kubernetes
gradually
6. Tool to learn new language
Source : https://assist-software.net/blog/pros-and-cons-serverless-computing-faas-comparison-aws-lambda-vs-azure-functions-vs-google
Slide 8
Slide 8 text
3 tinker board(2GB ram each )
300 euro in total(switch, power,etc)
Elixir, golang, python for ML
Slide 9
Slide 9 text
1. Community provided containers(OCR, QR code, Sentiment analysis)
2. GraphQL + Neo4j on OpenFaaS
3. S2
DEMO
Slide 10
Slide 10 text
Lab 1 - Prepare for OpenFaaS
Lab 2 - Test things out
Lab 3 - Introduction to Functions
Lab 4 - Go deeper with functions
Lab 5 - Create a Gitbot
Workshop plan (local machine)
Lab 6 - HTML for your functions
Lab 7 - Asynchronous Functions
Lab 8 - Advanced Feature - Timeouts
Lab 9 - Advanced Feature - Auto-scaling
Lab 10 - Advanced Feature - Secrets
Source: https://github.com/openfaas/workshop
Slide 11
Slide 11 text
Q&A
For chaining functions, please read this doc
https://github.com/openfaas/faas/blob/master/guide/chaining_functions.md
For async call(using a que like NATS),
https://github.com/openfaas/faas/blob/1aa6270fcc274cc36d90e0a9e4caa3eb
71912ae0/guide/asynchronous.md#call-a-function