Lustre Support Training in Copenhagen, Denmark

This course provides participants with the knowledge and hands-on experience to support, maintain, and troubleshoot Lustre parallel file systems. It covers Lustre architecture, installation, configuration, monitoring, and common operational issues. Participants will learn how to provide effective support for Lustre deployments in HPC (High-Performance Computing) environments, ensuring system stability, scalability, and performance.

Introduction to Lustre

• Overview of Lustre and its role in HPC

• Lustre architecture and components: MDS, OSS, MDT, OST, Clients

• Lustre networking basics (LNET, Infiniband, TCP/IP)

• Support workflows for Lustre environments

Installation & Configuration

• Lustre software stack and dependencies

• Setting up Lustre Metadata Server (MDS) and Object Storage Servers (OSS)

• Configuring clients for Lustre access

• Basic administration commands and tools

Operations & Monitoring

• Managing filesystems (mkfs.lustre, mount.lustre, tunables)

• Monitoring tools (Lustre utilities, LMT, Grafana/Prometheus integrations)

• Logs and error analysis

• User management & quotas in Lustre

Troubleshooting & Support

• Common Lustre issues and resolutions

• Debugging client/server communication problems

• Handling metadata and object storage failures

• Troubleshooting performance bottlenecks

• Backup, recovery, and failover strategies

Advanced Support & Best Practices

• Patching and upgrading Lustre systems

• Performance tuning parameters (RPC, caching, striping)

• Scaling Lustre for large clusters

• Security considerations in Lustre environments

Hands-on Exercises

Summary and Conclusion