Loading video player...
In this hands-on demo, we’ll see how Kubernetes keeps your machine learning app healthy even when it takes forever to start. You’ll learn how liveness, readiness, and startup probes work together to create self-healing, resilient, and production-ready workloads. Using a real FastAPI machine learning app, we’ll simulate model loading delays, containerize it, deploy it on Kubernetes, and watch how each probe kicks in to keep it alive. What you’ll learn: - How Kubernetes probes actually work - The difference between liveness, readiness, and startup probes - Why probe tuning is critical for real-world ML apps - How to write and deploy a complete YAML manifest - How to debug and fix probe-related issues ▬▬▬▬▬▬▬ Timestamps ⏰ ▬▬▬▬▬▬▬ 00:00 - Intro: Why Kubernetes Probes Matter (Real-World Use Case) 00:36 - How Startup, Readiness & Liveness Probes Work Together 01:59 - Simulating a Machine Learning App with FastAPI 03:19 - Containerizing the ML App with a Python Containerfile 04:40 - Defining All Three Probes in the Kubernetes Pod Manifest 05:41 - Deploying and Observing Pod Behaviour in Real Time 08:10 - Key Takeaways: Probe Tuning for Real Apps For 1:1 mentorship, consultations, or career guidance, you can book a session with me here: https://topmate.io/nikhil_kumar811/ 🔗 Watch Related Playlists: - Kubernetes: https://www.youtube.com/playlist?list=PL-K2rw28HIwZVMo9CtbV0wDu548SN0h9Y - Github Actions: https://www.youtube.com/playlist?list=PL-K2rw28HIwYfq7SqYnBzAxlUhcYP7ldM - Ansible: https://www.youtube.com/playlist?list=PL-K2rw28HIwaavCXTYEWF4mP431KmKtEY - AWX: https://www.youtube.com/playlist?list=PL-K2rw28HIwbTtijpBMrOaHdnWGXdOkYa - AI: https://www.youtube.com/playlist?list=PL-K2rw28HIwaSvmI8oFeSQDl4cVTdxaGQ ▬▬▬▬▬▬ Connect with me 👋 ▬▬▬▬▬▬ LinkedIn: https://www.linkedin.com/in/kumar-nikhil811/ Website: https://techinik.com Medium: https://medium.com/@kumarnikhil811 #kubernetes #cncf #devops #fastapi #cloudnative #machinelearning #ai