GitOpsCon North America 2022

Stateful workloads present a heavy challenge for platform operators to secure, backup, and provide for their recovery in the event of "black swan" or cluster-ending catastrophes. Developing a strategy for CSI volumes and managing snapshots should be a priority for cluster operators who otherwise may leave their important data vulnerable to ransomware attacks or other threats; what mitigations can be applied in the event of a full compromise? With GitOps, managing recovery of STATELESS workloads is made easy, almost effortless. It hasn't been shown how that level of convenience is replicated for STATEFUL workloads and DR on persistent volumes. Cluster operators may already know enough about CSI to understand that PVCs are backed by PVs and have a ReclaimPolicy that may be set to Retain. How does one manage this persistence with GitOps? How to manage snapshots? What exactly does snapshot restore executed via GitOps approach look like? Let’s endeavour now to offer clear prescriptive guidance about effective GitOps strategies for managing disaster recovery. What are some strategic approaches to apply to bring our risks back within tolerance? We will find out more in this talk.