Benny Bauer - Python in The Serverless Era

Serverless is the latest phase in the evolution of cloud development. Its building blocks are functions, a bunch of stateless “nano-services”, that can scale automatically and charged only when used. It enables teams to focus more on development while having fully managed servers. In this talk I'll cover the Serverless Architectures practices, use cases, tooling and the role python plays in it.

https://us.pycon.org/2017/schedule/presentation/469/

Transcript

1. Python in The Serverless Era Photo by Zack Spear Icon by Serverless
2. ◉ Chief Architect @ ◉ Ex- : Cloud Architect @ AutoCAD ◉ Find me at @benikbauer I am Benny Bauer Hello!
3. What is Serverless Dev SaaS BaaS FaaS
4. What is FaaS - Function as a Service Fully-managed compute Provisioning, patching, scaling, monitoring, logging are provided out-of-the-box Deploy your code Just package and upload the code Pay for actual usage Getting charged only upon code execution, per 100ms 100% UTILIZATION LESS OPS
5. How it works Deploy your code Define triggers Code execution EVENT-DRIVEN AUTO SCALING + AVAILABILITY ◉ HTTP requests ◉ Storage (e.g. file upload) ◉ DB (e.g. row insert) ◉ Scheduled tasks ◉ Messaging ◉ Many many more...
6. Other - Scheduled tasks - Distributed compute API - Web backend - Mobile backend - Bot Use cases Data processing - IoT - Streams (analytics, logging, etc) - Files (images, text, etc.) Operations - CI - Policy enforcement - Provisioning
7. Other - Scheduled tasks - Distributed compute API - Web backend - Mobile backend - Bot Use cases Data processing - IoT - Streams (analytics, logging, etc) - Files (images, text, etc.) Operations - CI - Policy enforcement - Provisioning
8. Things to be aware of Stateless Instances are ephemeral. Store state on client-side, cache or db. Cold start Latency (< 2 sec) when container is cold (upon first run or inactivity). Vendor lock-in Integrations with other services are the real bait. Granularity Unit of deployment is nanoservice/function. Costs Cost-effective up to a certain point. Limitations Execution time is 5 min. Payload & disk sizes are limited.
9. Let’s get to business! Python & FaaS
10. Cloud providers Python 2.7 & 3.6 are supported by: • AWS • OpenWhisk • Microsoft Azure - experimental, still not GA
11. Configuration - Event binding - Resources definition - Security roles definition The need for frameworks Deployment - Package - Upload - Rollback
12. Frameworks Chalice
13. Serverless Framework
14. Zappa One function only. Serves as a WSGI server WSGI
15. Unique features: • Global deployment • “Keep warm” functionality • SSL certification • Support for AWS Lambda compatible python libraries (lambda-packages & Manylinux wheels) Zappa
16. • Python Serverless Microframework for AWS • Each endpoint is a separate function Chalice
17. Chalice $ pip install chalice $ chalice new-project helloworld && cd helloworld $ cat app.py from chalice import Chalice app = Chalice(app_name="helloworld") @app.route("/") def index(): return {"hello": "world"} $ chalice deploy
18. Unique features: • Automatic IAM policy generation Chalice
19. Pywren
20. import pywren def scrape(url): # scrape it... return data wrenexec = pywren.default_executor() futures = wrenexec.map(scrape, [url,...]) results = pywren.get_all_results(futures) Credit: Sean Smith (https://blog.seanssmith.com/posts/pywren-web-scraping.html) Credit: Sean Smith (https://blog.seanssmith.com/posts/pywren-web-scraping.html)
21. Summary What was it all about?
22. Takeaways ◉ Serverless is a fast & cost-effective way to deliver many use cases with Python ◉ New frontiers for Python (frameworks, tooling, etc.)
23. Any questions? Thanks! Slides template by SlidesCarnival Find me at @benikbauer This work is licensed under a CC Attribution 4.0 International License.