Management of bare metal nodes is critical for High Performance Computing (HPC) clusters and supercomputers. The Ironic project provides a number of key features required for managing bare metal nodes in a cloud.
This talk covers various modifications and extensions made to Ironic by Cray to meet the requirements of HPC use cases across Cray's compute, storage and analytics product lines.
Additionally ORNL will cover the use case of a scientific cloud environment with a focus on leveraging HPC resources for high-impact research. Two enhancements to the HPC workflow are discussed: modifications to Ironic for supporting diskless boot, and progress on an application delivery method using Docker on Lustre.
Attendees will learn about a number of different use cases that Cray has for using Ironic in the context of HPC. We believe that many of these are not unique to Cray and will be of great interest to the OpenStack and HPC communities. We’ll discuss the changes that Cray has made to Ironic to support the use cases.
- Provisioning agent, based on the Fuel provisioning agent.
- Cinder integration for diskless nodes.
- Hardware discovery and inventory services.
- Support for multiple physical networks.
- Continuous integration testing of bare metal.
- Future scalability work in the pipeline.
Registration Opening Soon
Registration will open for you to purchase passes for the OpenStack Summit very soon!
We're excited to see you there!
