We want to ensure that Metal3 can scale to thousands of nodes and clusters. However, running tests with thousands of real servers is expensive and we don’t have access to any such large environment in the project. So instead we have been focusing on faking the hardware while trying to...
Running on bare metal has both benefits and drawbacks. You can get the best performance possible out of the hardware, but it can also be quite expensive and maybe not necessary for all workloads. Perhaps a hybrid cluster could give you the best of both? Raw power for the workload...
Metal3 project has introduced pivoting in its CI workflow. The motivation for pivoting is to move all the objects from the ephemeral/management cluster to a target cluster. This blog post will briefly introduce the concept of pivoting and the impact it has on the overall CI workflow. For the rest...
As a part of developing the Cluster API Provider Metal3 (CAPM3) v1alpha4 release, the Metal3 crew introduced a new project: its own IP Address Manager. This blog post will go through the motivations behind such a project, the features that it brings, its use in Metal3 and future work. What...
Metal3 supports multiple types of images for deployment, the most popular being QCOW2. We have recently added support for a feature of Ironic that improves deployments on constrained environments, raw image streaming. We’ll first dive into how Ironic deploys the images on the target hosts, and how raw image streaming...