Lightning Talk: How to Avoid a Kubernetes Doom Loop

This lightning talk is a tale of what happens when automation becomes the problem.

A single cluster running 16K Argo Workflows across more than 165 nodes, suddenly begins to melt down.

One misconfigured pod starts to automate pod failures on a huge scale. Overloading a cluster is a costly recipe for failure. In this short talk, hear from bitter experience and find out why best practices provide the best basis for an easier life as a CRE.