Choose timezone
Your profile timezone:
CERN STEAM Academy 2026 - Timetable & Lecture Programme
Modern Storage Systems: From Single Bits to Distributed Infrastructures
by ,
→
Europe/Zurich
593/R-010 - Salle 11 (CERN)
Description
Abstract
Every time you open a file something slightly improbable happens. A request falls through a dozen layers of software, a spinning disc or a grid of leaking charges is persuaded to give back exactly what you left there, and, if you are lucky, nobody lies to you along the way.
This course follows that journey. We begin at the platter and the flash cell, climb up through the Linux I/O stack, and end at the Worldwide LHC Computing Grid, where the file you asked for may well be sitting on a machine on the other side of the world. Along the way one lesson keeps returning: there are no free choices. Every mechanism that makes storage faster, cheaper or safer sends the bill somewhere else, and the only way to choose well is to know what your workload actually does.
No prior storage experience is assumed. Bring a laptop: three hands-on sessions have you build a small distributed object store, measure your own machine, and watch a file corrupt itself while every tool reports success.
Lecturer Bio
Gianmaria is a Software Engineer at CERN in the IT Storage group, where he works on the development and maintenance of EOS, the large-scale distributed storage system that manages the massive data volumes produced by the LHC experiments. His work spans the full life of exabyte-scale storage, from the design of new features to keeping a system serving petabytes a day reliable in production.
Tutor's Bio
Octavian is a Software Engineer at CERN in the IT-SD-PDS group, where he manages EOS ATLAS and EOS AMS, the storage systems holding physics data for two of the LHC experiments. Combined with the storage backing the rest of the experiments, that's over an exabyte of data underpinning CERN's physics research. He interacts with the EOS codebase itself, both as a developer and operator. His work also covers the full lifecycle of that storage, from the hardware stack to on-site node investigations and firmware updates, to building the automation that keeps the fleet reliable day to day.