Scaling Python Web Applications With Kubernetes and Karpenter
The Real Python Podcast
Exploring AI Model Costs and Community Engagement
This chapter examines the challenges and costs of running AI models, comparing local instances with hosted services. It also highlights upcoming Python meetups, community engagement, and expresses appreciation for a technology sponsor's support.
What goes into scaling a web application today? What are resources for learning and practicing DevOps skills? This week on the show, Calvin Hendryx-Parker is back to discuss the tools and infrastructure for autoscaling web applications with Kubernetes and Karpenter.
Calvin is the co-founder and CTO of Six Feet Up, a Python and AI consultancy. He shares how they recently helped a client scale a web application that employs video, audio, and chat sessions. We dig deep into the tooling behind modern Kubernetes systems management and performance monitoring.
Calvin shares a project bootstrap tool for streamlining the development and deployment of a web application. The tool includes a complete blueprint for the infrastructure needed to get started.
We also dig into a collection of coding tools Calvin has been experimenting with. We discuss his recent IndyPy presentation, “Battle of the Bots,” which put several AI code assistants through their paces.
This episode is sponsored by AMD.
Course Spotlight: First Steps With LangChain
Large language models (LLMs) have taken the world by storm. In this step-by-step video course, you’ll learn to use the LangChain library to build LLM-assisted applications.
Topics:
- 00:00:00 – Introduction
- 00:02:23 – Scaling a Django project using Kubernetes
- 00:05:35 – Elastic Kubernets Service (EKS)
- 00:09:10 – Kubernetes terminology and improvements in tooling
- 00:11:29 – The Control Plane and the API
- 00:14:06 – Video Course Spotlight
- 00:15:11 – scaf: providing DevOps engineers a blueprint for new projects
- 00:17:21 – What have been the benefits of scaf for internal teams?
- 00:20:18 – How do you identify and reproduce scaling issues?
- 00:22:44 – Dealing with IP address scaling
- 00:26:03 – Why use other observability tools beyond AWS internal ones?
- 00:29:22 – Other lessons learned and moving toward refactoring code
- 00:33:53 – Scaling a voice-based LLM application
- 00:35:35 – Sponsor: AMD
- 00:36:11 – Looking at limitations and bottlenecks
- 00:38:08 – Configuring a Kubernetes operator to act on itself
- 00:39:47 – What project components are within a pod of containers?
- 00:42:31 – Budgeting for scale using Karpenter
- 00:43:58 – Tools for running containers locally
- 00:46:01 – Are containers still a primary development tool for you?
- 00:50:58 – Resources for learning DevOps and Kubernetes
- 00:52:54 – Conferences and talks
- 00:53:56 – Battle of the Bots: comparing coding agents
- 00:55:15 – What are you excited about in the world of Python?
- 00:56:20 – What do you want to learn next?
- 01:02:42 – What’s the best way for people to follow your work online?
- 01:03:33 – Thanks and goodbye
Show Links:
- Six Feet Up - Python and AI for Good, Custom Software Development
- Kubernetes - Tutorials
- Managed Kubernetes Service - Amazon EKS - AWS
- Karpenter
- Kustomize - Kubernetes native configuration management
- Kubernetes Components - Control Plane Components
- Continuous Integration and Deployment for Python With GitHub Actions
- Argo CD
- scaf: Provides developers and DevOps engineers with a complete blueprint for a new project
- Streamline the Dev Experience with Kubernetes and Scaf™
- Scaf™ — Six Feet Up
- Scaf: Complete blueprint for new Python Kubernetes projects - Talk Python To Me Podcast E496
- kind
- Locust - A modern load testing framework
- Grafana: The open and composable observability platform
- Grafana Loki OSS - Log aggregation system
- Prometheus - client_python
- Elastic network interfaces - Amazon Elastic Compute Cloud
- eks-node-viewer: EKS Node Viewer
- k9s: 🐶 Kubernetes CLI To Manage Your Clusters In Style!
- OrbStack · Fast, light, simple Docker & Linux
- NixOS Wiki - Python
- TechWorld with Nana - YouTube
- Python: The Documentary [OFFICIAL TRAILER] - YouTube
- Calvin Hendryx-Parker - LinkedIn
Conferences and Meetups:
- All Things Open 2025 - All Things Open
- All Things Open AI Conference
- All Things Open AI 2025 - AI Builders Track - YouTube
- Rolling out Enterprise AI: Tools, Insights, & Team Empowerment - Calvin Hendryx-Parker, Six Feet Up - YouTube
- PyCon US 2025 - PyCon US 2025
- PyOhio 2025
- IndyPy Events
AI Coding Tools:
- Battle of the Bots - Developer Tools Showdown - YouTube
- Aider - AI Pair Programming in Your Terminal
- codename goose
- An entirely open-source AI code assistant inside your editor - Ollama Blog
- Devstral - Mistral AI
- 10 LLM Observability Tools to Know in 2025 - Coralogix
- How often do LLMs snitch? Recreating Theo’s SnitchBench with LLM
- The lethal trifecta for AI agents: private data, untrusted content, and external communication
- vllm-proxy: Proxy for vLLM enabling multi-model operation, cache-aware routing, and load balancing.
Level up your Python skills with our expert-led courses:
The AI-powered Podcast Player
