Glossary Term

ROT data

What is ROT data?

ROT data is the digital clutter that accumulates in every organization: duplicate files nobody needs, outdated records nobody uses, and trivial stuff - personal photos, throwaway notes - that never had business value in the first place. It takes up storage, inflates costs, and gets in the way of finding what actually matters.

The acronym breaks into three categories. Redundant data is duplicates: copies of the exact same file and file version living on someone’s desktop, a shared server, and two cloud folders because it got copied during collaboration or backups. Obsolete data is information that has aged out – files from employees who left years ago, superseded document versions, records that no longer apply. Obsolete data can also apply to a wide variety of temporary files, such as BCL files and temporary BAM files created during scientific research. Trivial data is the rest: personal photos, casual emails, temp files that could vanish tomorrow and nobody would notice. Various industry estimates put ROT at anywhere from 33% to 85% of organizational content, and roughly half of all stored data goes unaccessed for two years or more.

What’s key is that for HPC environments, research institutions, and enterprises sitting on billions of files, ROT dead weight pushes storage costs up 20–30% a year while widening the security attack surface and making compliance with GDPR, HIPAA, and CCPA harder than it needs to be.

At extreme scale, you cannot clean this up manually. You need metadata-driven visibility and automation. Starfish Storage’s platform tackles ROT through its Unstructured Data Catalog system that tracks file age, access, duplication, and content across heterogeneous storage; departmental NAS, petabyte-scale Lustre and GPFS, Weka, VAST and all other file systems. The In-Depth Browser Analytics feature lets both admins and end users spot candidates for archiving or deletion within any directory without running separate reports, and Starfish Zones give users self-service access so they do not have to wait on IT for every cleanup request. The platform enriches metadata across 100+ file formats and supports asynchronous searches across billions of files, and job execution on those results, enabling organizations to systematically find and remove ROT, free up premium storage tiers for active data, cut backup costs, and get datasets into better shape for AI/ML workloads. For HPC centers operating at exabyte scale – think El Capitan and other Top500 systems – routine metadata-driven ROT cleanup is basic infrastructure hygiene. Skip it and you are paying more than you should for storage you are not really using.