This course is a workshop-focused, hands-on class designed to give system administrators the knowledge and confidence required to implement and maintain IBM Platform HPC in their HPC environment. Students will gain a solid and fundamental understanding of the product's features, including how the product is implemented, configured, and administered. The workshops provide hands-on experience with the material presented during class.
Students should have:
- At least some Linux/Unix command-line experience
No previous experience with IBM Platform products is assumed or required.
- Concepts and Terminology
- What's included in IBM Platform HPC
- Resource Management
- Managing Cluster Resources Command Line Tools
- Host Provisioning Details
- Workload Manager Overview
- Using the Workload Manager
- Administering the Workload Manager
- Workload Manager Scheduling Policies
- Running Parallel jobs
- Application Integration
- Monitoring, Alerting and Reporting
- Support and Troubleshooting
- The Building Kit
- Dynamic Multiboot
- Custom Networks
- NIC Bonding
- Kusu RC script framework
- The Kusu Database
The target audience for this course are System administrators who will implement the HPC cluster and who will be responsible for managing the environment.