Learn · CERN

Talk22 min

Ceph Scaling Strategies for Machine Learning Workloads

Radomir Michal Wasowski · CERN

Watch At CERN

About

Ceph is a software-defined solution designed to manage large-scale data stores. It ensures high availability and concurrent access while providing a POSIX-compliant file system (CephFS), making it an attractive option for serving distributed machine learning workflows. This project focuses on optimising CephFS resource usage to improve overall storage performance and reduce the bottleneck of data access in such use-cases. We will present our testbed and benchmarks of various CephFS configurations to examine possible optimizations from cluster-administrator and user perspectives.

0:00:00 Slide 1

0:00:00 Slide 2

0:00:14 Slide 3

0:00:30 Slide 4

0:00:52 Slide 5

0:01:55 Slide 6

0:02:51 Slide 7

0:03:51 Slide 8

0:04:51 Slide 9

0:05:18 Slide 10

0:05:58 Slide 11

The talk

  1. Ceph Scaling Strategies for Machine Learning Workloads 2026-03-05 · 22 min · CERN's player

Credits and licence

Institution
CERN
Speaker
Radomir Michal Wasowski
Year
2026
Language
English
Subjects
Computer science
Source
CERN
Licence
CERN, as CERN states it. CERN's terms
Attribution
© CERN. CERN provides its audiovisual media free of charge for educational and informational use, credited to CERN; this lecture is shown in CERN's own player.
Who can watch
CERN gives no rating. Invisible Network shows it to Tweens profiles and older, judged from what the source says about it. How courses are placed