An HPC (High-Performance Computing) cluster is a collection of networked machines (nodes) that work together to perform large-scale computations in parallel. These clusters typically consist of:
- Management Nodes (Master / Worker) – Control and orchestration
- Login Nodes – User access points
- Compute Nodes (Diskless) – Execute workloads
All nodes are interconnected through a high-speed network to enable efficient communication and data sharing.
In diskless HPC environments, compute and login nodes require a mechanism to boot over the network. This is achieved using iSCSI-based Boot Content Projection.
iSCSI (Internet Small Computer Systems Interface) is a block-level storage protocol that transmits SCSI commands over TCP/IP networks.
iSCSI SBPS (Scalable Boot Projection Service) enables:
- Remote booting of diskless nodes
- Projection of boot images such as:
- Root Filesystem (rootfs)
- Pre-Execution Environment (PXE/PE) images
- Centralized storage management for scalable clusters
To develop a robust CLI-based management utility that:
- Runs on the Master Node
- Retrieves iSCSI SBPS-related data from all target nodes
- Eliminates the need for manual login into each node
- Provides centralized visibility into storage and boot configuration
- Cluster-wide iSCSI target discovery
- Automated iSCSI related data collection across nodes
- Lightweight and scriptable CLI interface
For detailed guides, please refer to the following documentation files:
- Setup Guide — Comprehensive guide to VM installation, DNS configuration, passwordless root SSH setup,
pdshsetup, and Kubernetes node provisioning. - CLI Reference — Complete reference for the
iscsicommand-line utility, including available subcommands, arguments, syntax, and output placeholders.