by Datadog


With 23 releases since 2015, and more than 3,000 contributors, Kubernetes is the most successful open source container orchestrator. And with all major public clouds now having a supported and managed solution, migrating workloads to Kubernetes is easier than ever.

But although starting with Kubernetes is approachable, a set of common problems arise when scaling from a few dozen nodes, to hundreds of them across multiple availability zones, and from several services to running all your services as Kubernetes workloads.

In this roundtable we’ll discuss the challenges that several companies running their business on top of large Kubernetes clusters faced when scaling their clusters. The participants will include members from SRE teams of these companies who are in charge of maintaining these clusters and to provide them to the rest of the engineering teams. They will share their stories from the trenches, including bugs and limitations they found along the way and how they overcame them.