LUG 2026

Lustre Users Group 2026

April 27 – 29, 2026

Indianapolis, IN

 

LUG is the conference for all things related to Lustre shared parallel storage.  LUG 2026 was held in Indianapolis, Indiana, USA April 27-29, 2026. LUG 2026 included new features/improvements in Lustre including the most-recent release, Lustre 2.17 with Hybrid IO, Dynamic LNet NID Configuration and Nodemap enhancements.

There were also presentations on new and innovative uses of the Lustre File System in High Performance Computing (HPC) furthering key research and AI model training, some best practices, and perhaps something to surprise you.

Day 1: April 28, 2026

Time Title Presenter / Affiliation Slides
8:00 Welcome/Breakfast
9:15 Welcome Talk OpenSFS and Indiana University
9:30 Lustre Community update Peter Jones, The Lustre Collective Slides
10:00 Break (30min)
10:30 Lustre 2.18 and Beyond Andreas Dilger, The Lustre Collective Slides
11:30 Upstream efforts progress James Simmons, ORNL Slides
12:00 Lunch (1h)
13:00 Sponsor talk DDN
13:15 Lustre In The Works Sebastien Buisson, DDN Slides
13:45 Lustre transition to folio i/o Shaun Tancheff, HPE Slides
14:15 Status of Lustre’s Hybrid I/O Marc-Andre Vef, DDN Slides
14:45 Break (30min)
15:15 Lustre Erasure Coding Patrick Farrell, The Lustre Collective Slides
15:45 Metadata redundancy for Lustre Oleg Drokin and Lai Siyao, DDN Slides
16:15 OpenSFS Update OpenSFS Board
17:00 End of program for day 1
17:30 Social Event starts

Day 2: April 29, 2026

Time Title Presenter / Affiliation Slides
8:00 Welcome/Breakfast
9:00 The MLPerf Storage Benchmark and Lustre: Overview and Best Practices Ellis Wilson, Microsoft Slides
9:30 Enhancing Lustre for Modern AI with Disaggregated Storage and ML-Tuned Caching Policies Sergei Platonov, Xinnor Slides
10:00 Break (30min)
10:30 LNet observability and monitoring Timothy Day, The Lustre Collective Slides
11:00 Data Management Practices on Large Multi-Tier Lustre Filesystems Richard Mohr, Ethan Frenza, Brad Gipson and Jesse Hanley, ORNL Slides
11:30 Tips and Tricks for Empowering LLMs with Lustre Knowledge Ellis Wilson, Microsoft Slides
12:00 Lunch (1h)
13:00 Sponsor talk AWS
13:15 Practical Multi-Tenancy for HPC/Cloud in 2026 Robert Read and Matt Raso-Barnett, The Lustre Collective Slides
13:45 Architectural Lessons and Operational Case Studies from OCI’s Managed Lustre Service Aboo Valappil, Sonia Sharma and Vinod Johnson, Oracle Slides
14:15 High-Performance Cloud-Native Lustre on Object Storage with Active-Active HA Supramani Sammandam, Zettalane Systems Slides
14:45 Break (30min)
15:15 Lustre on Kubernetes: Leveraging Advanced Lustre Features for AI/HPC Workloads via CSI Integration Shuichi Ihara, DDN Slides
15:45 HALO: A New System for Lustre HA Management Thomas Bertschinger, LANL Slides
16:15 Mount IAM security model with lustre Di Wang, Google Slides
16:45 Empowering HPC and AI Workloads with optimized Lustre reliability and performance Duncan Vogel and Zanhua Huang, AWS Slides
17:15 Closing Remarks OpenSFS
17:30 End of LUG 2026

 

Thank You to our Sponsors! 

DataDirect Networks (DDN)

Platinum Sponsor

Amazon Web Services (AWS)

Platinum Sponsor

Aeon Computing

Gold Sponsor

Google Cloud

Silver Sponsor

The Lustre Collective

Bronze Sponsor

Files powered by

 

 

Conference by OpenSFS and Indiana University Conferences