Running Apps in Kubernetes: I Broke Them So You Don't Have To

So, you've decided to deploy your application on Kubernetes—congratulations! Now, let me save you from making all the same mistakes I did. In this session, we'll explore the most common ways people (myself included) mess up their first Kubernetes apps and, more importantly, how to fix them without becoming a certified Kubernetes guru.

This is Kubernetes 101—but with a twist. Instead of focusing on how to set up Kubernetes, we'll focus on both code-level changes and simple manifest/config updates that help you actually take advantage of Kubernetes' features to make your applications more resilient, scalable, and easier to manage. We'll cover:

• Probes Done Right: Why liveness and readiness probes matter, and how to implement them correctly to avoid cascading failures.
• Resource Requests and Limits: How to stop starving your pods—or wasting resources.
• Horizontal Pod Autoscaling (HPA): Scaling like you mean it. Why replicas and auto-scaling are essential to justify using Kubernetes at all.
• Metrics That Matter: Leveraging Prometheus (or alternatives) to track performance and ensure visibility into your workloads.
• Keep Traffic Local: Avoid unnecessary egress traffic and make the most of Kubernetes networking.
• Jobs and CronJobs: Scaling batch processing correctly instead of stuffing everything into a single pod.
• Graceful Terminations: Avoid data loss and downtime by handling SIGTERM signals the right way.
• We'll wrap up with a checklist of "low-hanging fruit" improvements you can make today to get the most out of Kubernetes—or decide if you'd be better off using something simpler like ECS or Lambda.

Key Takeaways:

• Learn the critical Kubernetes features you should be using.
• Discover common pitfalls and how to avoid them.
• Assess whether Kubernetes is actually the right fit for your application.

This session is perfect for developers and DevOps engineers who are new to Kubernetes or struggling to make their applications production-ready. You'll leave with practical tips to improve your manifests, configurations, and code