From Pre-Population to Disasters: Managing and Protecting VM State in KubeVirt

Explore the intricacies of managing and protecting virtual machine states in a Kubernetes environment through this comprehensive 28-minute conference talk by Michael Henriksen from Red Hat. Dive into the world of KubeVirt and learn how it enables running traditional VM workloads within Kubernetes clusters. Discover the KubeVirt storage layer's role in bridging QEMU/KVM Virtual Machines with K8s storage primitives, offering a feature-rich API for VM data management throughout its lifecycle. Gain insights into creating new VMs with pre-populated disks using "golden images" and implementing safe snapshot and restore procedures for running VMs. Examine disaster recovery workflows enabled by the VirtualMachineExport API and Velero integration. Acquire a high-level understanding of KubeVirt storage APIs and architecture, empowering you to make meaningful contributions to the field. The talk covers various topics, including persistent storage options, DataVolume API, DIY golden image provisioning, VirtualMachineSnapshot and VirtualMachineRestore APIs, and future initiatives like Volume Populator support.