Empowering AWS DevOps with Python and Machine Learning

Explore a conference talk that delves into optimizing AWS DevOps practices using Python and machine learning techniques. Learn how to tackle scaling challenges in cloud infrastructure by leveraging data analysis and simulation. Discover methods for retrieving AWS data with pandas, analyzing host performance, and creating scaling simulators. Gain insights into traffic shape generation, local experimentation, and parameter optimization. Compare simulation results with real-life production testing, and understand the practical implications for improving cloud infrastructure efficiency. Acquire valuable knowledge on integrating Python and machine learning to enhance AWS DevOps workflows and decision-making processes.

intro
preamble
intro
agenda
scaling is hard
problem
context
approach
load test one host
retrieve data from aws to pandas
analyse one host performance
we can do better
another thing - latency
approaches
create a scaling simulator
create a traffic shape generator
and run a first local experiment
what does it mean?
first attempt
find the best parameters
and the winner is
which actually looks better
testing new parameters in production
simulation versus real life
conclusion
code
about gustavo