Exploring OpenShift Virtualization with Ansible

[DamianB]

I'd like to ask (as a RHEL beginner): What can I actually do with OpenShift Virtualization and Ansible, and what are the key benefits?

Right now, I’m testing different versions of CAE software, along with their modules. Using some bash scripts and Ansible playbooks, but only on a single RHEL VM in libvirt.

From what I understand, OpenShift Virtualization could take this to the next level. So, when do we ramp up activity in this group to dive deeper?

[Chetan_Tiwary_ · Accepted solution]

@DamianB    https://www.redhat.com/en/blog/ansible-automation-platform-openshift-virtualization-multi-cluster-en... 

The perfect solution for your CAE testing workflow is the combination of OpenShift Virtualization (KubeVirt) and Ansible. KubeVirt allows you to run your traditional VMs as native K8s objects within OCP, giving you modern orchestration for your legacy workloads. Ansible then automates the entire process. Forget your old setup of a single RHEL VM and manual bash scripts. With this new stack, you can instantly scale your testing to 100s of VMs distributed across the cluster, using Ansible playbooks to deploy tailored VM fleets (e.g., v1 test environments vs. v2 test environments) and manage them with built in autoscaling and load balancing. Setting up a test is faster than ever; Golden VM images let you spin up a clean environment with the required CAE software and modules in seconds. Finally, you can leverage PersistentVolumes and CSI storage to create shared, versioned datasets, complete with snapshots, which is a significant improvement over isolated local storage.

[DamianB]

Absolutely agree, this is exactly the setup my CAE workflows need!

Perfect for parallel optimization studies, DOE's and generating datasets for AI surrogates. It's the kind of clarity that makes you wonder why we didn't do this sooner. Winter is indeed the best time to dig in!