icaoberg icaoberg

A reference glossary of the key components that make up a Lustre parallel file system—MDS, MDT, OSS, OST, MGS, clients, LNet, and the terms you need to read a Lustre layout.

A definition of Lustre, how its architecture achieves massive parallel bandwidth, and how it compares to other file systems common in HPC—GPFS/Spectrum Scale, BeeGFS, NFS, and Ceph.

A formal definition of a file system, an intuitive TL;DR, and a look at parallel file systems with a focus on Lustre.

An introduction to HPC and how it relates to — and differs from — distributed systems.

A quick overview of SLURM, the workload manager used on most HPC clusters.

A quick overview of Apptainer (formerly Singularity), the container runtime built for HPC environments.

Why CLI apps remain essential and how Singularity makes them portable on HPC systems.

Data curation is the active, ongoing management of data to ensure it remains accurate, findable, and useful long after it is collected. This post explains what data curation is, what data curators do, why the role is essential in modern research, and how PSC applies it to national-scale scientific data infrastructure.

A practical introduction to SLURM job scheduling on PSC Bridges-2, covering sbatch scripts and interactive srun sessions for both CPU and GPU partitions.

FAIR is a set of guiding principles for scientific data management that has become a global standard. This post explains what each principle means, why it matters in modern science, and what it demands from researchers who have been doing this for years — especially around code, workflows, and reproducible computing.