Stéphane Thiell

@sthiell.bsky.social

I do HPC storage at Stanford and always monitor channel 16 ⛵

Just expanded Sherlock’s /scratch to 15 PB of all-flash Lustre at Stanford Research Computing. Built in-house on community Lustre with our own backported patches, OSTs added live. No vendor black box. No appliance magic. Just open source and a couple people who know their way around it! 🧱🔧⚡️

Scratch that: Sherlock now has 15 PB of flash - Sherlock changelog

We are excited to announce a major expansion of Sherlock's /scratch file system. We are adding 5 PB of full-flash storage, bringing the total capacity from 10 PB to 15 PB. This investment directly add...

news.sherlock.stanford.edu

ClusterShell 1.9.3 is now available in EPEL and Debian. Not using clustershell groups on your #HPC cluster yet?! Check out the new bash completion feature! Demo recorded on Sherlock at @stanford-rc.bsky.social with ~1,900 compute nodes and many group sources! asciinema.org/a/699526

clustershell bash completion (v1.9.3)

This short recording demonstrates the bash completion feature available in ClusterShell 1.9.3, showcasing its benefits when using the clush and cluset command-line tools.

asciinema.org

Fun fact: the Georgia Aquarium (nonprofit), next to the Congress center is the largest in the U.S. and the only one that houses whale sharks. I went on Sunday and it was worth it. Just in case you need a break from SC24 this week… 🦈

Bild

I always enjoy an update from JD Maloney (NCSA), but even more when it is about using S3 for Archival Storage, something we are deploying at scale at Stanford this year (using MinIO server and Lustre/HSM!)

BildBildBild

Filesystem upgrade complete! Stanford cares about HPC I/O! The Sherlock cluster has now ~10PB of full flash Lustre scratch storage at 400 GB/s, to support a wide range of research jobs on large datasets! Fully built in-house!

Racks of Fir Storage at the Stanford Research Computing Facility (SRCF)df command showing 9.8PB /scratchArtist photo of an OSS pair of Fir with HDR-200