Standalone yet at scale: Ironic take on High Performance Computing
summary
Los Alamos prototypes a standalone Ironic deployment that provisions bare metal at scale, aiming to turn weeks into minutes, manage mixed fleets from 32 to >20,000 nodes, and serve both disk-full and diskless HPC workloads plus tens of petabytes of filesystems. It pairs with Ansible to replace legacy CFEngine and Perseus, uses conductor groups to scale horizontally, supports IPMI and Redfish, and demonstrates fast HTTP-based image deployment and a working fastrack standby provisioning flow.
AI-written summary · about
chapters (5)
- 0:00Introduction to Ironic
- 3:44HPC flexibility challenges
- 7:15Ironic standalone benefits
- 9:23User perspective at LANL
- 15:00Summary and Q&A
related talks
16mA novel Cloud High Performance Computing infrastructure for user
27mFrom Supercomputing to Serving: A Case Study Delivering Cloud Native Foundation Models
47mCloud Native and Sustainable, Reproducible Scientific Computing
26mOpenHPC on Ansible
10mLightning Talk: Does Cloud Elasticity Pay off for HPC Workloads?
27mSupporting Security Sensitive Tenants in a Bare-Metal Cloud
28mHow to Deploy CF and Other Bosh Releases on Bare-Metal Environment
21mKeynote: CERN Experiences - Ricardo Rocha & Clenimar Filemon