Kubernetes Persistence: How to Run Stateful Workloads Without Losing Data

Welcome back, CNCF Kaduna!

After a brief hiatus, we are returning with a session that tackles one of the most critical challenges in modern infrastructure: State.

While Kubernetes was born for stateless workloads, the reality of modern engineering requires us to run databases, message queues, and complex data pipelines within our clusters. The stakes are high, get it wrong, and you lose data; get it right, and you have a truly resilient, automated platform.

In this session, Abdullahi Mustapha, a seasoned DevOps Engineer, will move beyond the basics to explore how we can manage data reliably and at scale. This is a "fluff-free" look at how to ensure your persistence layer is as robust as your compute.

What we will cover:

• The Persistence Problem: Why managing state in K8s is fundamentally different from stateless apps and why "Deployments" aren't enough.
• Storage Primitives: A technical look at Persistent Volumes (PVs), Claims (PVCs), and how to leverage dynamic Storage Classes.
• StatefulSets & Operators: A deep dive into pod ordinals and when to use native controllers versus third-party operators for complex databases.
• Real-World Reliability: Hard-earned strategies for backups, disaster recovery, and ensuring your data survives even when your nodes don't.

Whether you’re a DevOps engineer looking to migrate your first database to K8s or a developer curious about how your data actually lives in the cloud, this session is for you.

Agenda:

• **Community Welcome & Networking: ** 10 mins
• Technical Session: 45 mins
• Interactive Q&A / Architecture Review: 20 mins

Speaker Bio: Abdullahi Mustapha is a DevOps Engineer with extensive experience in DevOps and Platform Engineering. He specializes in building resilient, automated infrastructure and helping teams navigate the complexities of the cloud-native ecosystem.