AKS Survival Pack: what to consider before going all-in with AKS

on YouTube ↗01h 02m

" summary

Azure Kubernetes Service manages the control plane but leaves worker nodes, upgrades and hardening to the user, and neglecting basics causes real incidents: missing CPU/memory limits leads to evicted pods and cluster-wide outages, container images can be 13–14 GB images that take 10–15 minutes to pull, and one study finds 37% underutilized compute in clusters. Recommendations include Azure Policy for guardrails, GitOps with caution, and cost/security tooling such as kube-cost/OpenCost; demos and GitHub playbooks illustrate fixes.

AI-written summary · about

" chapters (9)

1. 0:00Introduction to AKS
2. 6:34Application readiness
3. 13:01Human and cultural factors
4. 16:07Developer experience
5. 21:26GitOps and deployment
6. 23:55Cost and sustainability
7. 32:45Security considerations
8. 44:02Operations and reliability
9. 51:04Q&A and closing

" related talks

• 33mManaged Kubernetes Service: Day Zero Survival PackKristina Devochko · KCD Ukraine 2023
• 32mMastering Cloud Kubernetes: Navigating Azure's Kubernetes Service MazeFlorin Loghiade · KCD Romania 2024
• 30mManaged Kubernetes Service Day Zero Survival PackKristina Devochko · KCD Amsterdam 2023
• 57mAzure Kubernetes Service (AKS) for BeginnersPrakash Tripathi · Full Stack Virtual Conference 2021
• 16mCracking the code: AI workloads at scale on AKSShon Harris, Yaroslav Meniailov · Microsoft Build 2024
• 56mAzure Kubernetes Services 101 for DevelopersJorge Arteiro · NDC Melbourne 2022
• 01h 01mDevOps with Azure Kubernetes ServiceJakob Ehn · NDC London 2021
• 44mAzure Linux: A container host OS for Azure Kubernetes Service (AKS) Q&AJim Perrin, Sudhanva Huruli, Sean McKenna, Tim Fish, Kellen Anderson, Olivia Al-Joundi · Microsoft Build 2023