Lustre User Group 2025: Difference between revisions

From Lustre Wiki
Jump to navigation Jump to search
 
(6 intermediate revisions by 2 users not shown)
Line 3: Line 3:


== LUG 2025 Agenda ==
== LUG 2025 Agenda ==
=== Tuesday, May 7 – Day 1 ===
=== Tuesday, Apr 1 – Day 1 ===
{| border=1 cellpadding=0
{| border=1 cellpadding=0


Line 13: Line 13:
|-
|-
|''' 09:30-10:00'''
|''' 09:30-10:00'''
|'''[[Media:LUG2025-Unraveling_the_Universe_with_HPC-Alvarez.pdf|Unraveling the Universe with High Performance Computing]]'''
|'''[[Media:LUG2025-Unraveling_the_Universe_with_HPC-Alvarez.pdf|Unraveling the Universe with High Performance Computing]]''' ([https://www.youtube.com/watch?v=0ot153D1BW0&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=18&pp=iAQB video])
:  Marcelo Alvarez, Kavli Institute for Particle Astrophysics and Cosmology
:  Marcelo Alvarez, Kavli Institute for Particle Astrophysics and Cosmology


|-
|-
|''' 10:00-10:30'''
|''' 10:00-10:30'''
|'''[[Media:LUG2025-Community_Release_Update-Jones.pdf|Lustre Community Update]]'''
|'''[[Media:LUG2025-Community_Release_Update-Jones.pdf|Lustre Community Update]]''' ([https://www.youtube.com/watch?v=fjkusGd8OZ8&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=16&pp=iAQB video])
:  Peter Jones, Whamcloud
:  Peter Jones, Whamcloud


|-
|-
|''' 11:00-11:30'''
|''' 11:00-11:30'''
|'''[[Media:LUG2025-Streamlining_AI_Checkpoints-Degremont.pdf|Streamlining AI Checkpoints: Automatic Data Migration to Object Store]]'''
|'''[[Media:LUG2025-Streamlining_AI_Checkpoints-Degremont.pdf|Streamlining AI Checkpoints: Automatic Data Migration to Object Store]]''' ([https://www.youtube.com/watch?v=9WaErzql9qg&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=17&pp=iAQB video])
:  Aurélien Degrémont, NVIDIA
:  Aurélien Degrémont, NVIDIA


|-
|-
|''' 11:30-12:00'''
|''' 11:30-12:00'''
|'''[[Media:LUG2025-Lustre_Upstreaming_Efforts-Simmons.pdf|Lustre Upstreaming Efforts]]'''
|'''[[Media:LUG2025-Lustre_Upstreaming_Efforts-Simmons.pdf|Lustre Upstreaming Efforts]]''' ([https://www.youtube.com/watch?v=nOPpWygOreI&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=15&pp=iAQB video])
 
:  James Simmons, Oak Ridge National Laboratory
:  James Simmons, Oak Ridge National Laboratory


|-
|-
|''' 13:00-13:15'''
|''' 13:00-13:15'''
|'''Sponsor Talk: [[Media:LUG2025-EXAScaler_Feature_Overview-Poddubnyy-Barnett.pdf|EXAScaler Feature Overview]]'''
|'''Sponsor Talk: [[Media:LUG2025-EXAScaler_Feature_Overview-Poddubnyy-Barnett.pdf|EXAScaler Feature Overview]]''' ([https://www.youtube.com/watch?v=q0rgj4Wn6EQ&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=14&pp=iAQB video])
:  Ivan Poddubnyy, DDN; Matt Rásó-Barnett, DDN
:  Ivan Poddubnyy, DDN; Matt Rásó-Barnett, DDN


|-
|-
|''' 13:15-13:45'''
|''' 13:15-13:45'''
|'''[[Media:LUG2025-ALCF_Site_Update-Kulyavtsev.pdf|ALCF Site Update]]'''
|'''[[Media:LUG2025-ALCF_Site_Update-Kulyavtsev.pdf|ALCF Site Update]]''' ([https://www.youtube.com/watch?v=YOjRpiggm9s&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=13&pp=iAQB video])
:  Alex Kulyavtsev, Argonne National Laboratory
:  Alex Kulyavtsev, Argonne National Laboratory


|-
|-
|''' 13:45-14:15'''
|''' 13:45-14:15'''
|'''[[Media:LUG2025-Customer_IO_DOS_Mitigation_NRS_TBF-Aguilar-Peter.pdf|Sandia Labs Lustre Filesystem Customer IO DOS Mitigation Using NRS Token Bucket Filters]]'''
|'''[[Media:LUG2025-Customer_IO_DOS_Mitigation_NRS_TBF-Aguilar-Peter.pdf|Sandia Labs Lustre Filesystem Customer IO DOS Mitigation Using NRS Token Bucket Filters]]''' ([https://www.youtube.com/watch?v=EoSmszy5zMQ&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=12&pp=iAQB video])
:  Michael Aguilar, Sandia National Laboratories; Jason Peter, Sandia National Laboratories
:  Michael Aguilar, Sandia National Laboratories; Jason Peter, Sandia National Laboratories


|-
|-
|''' 14:45-15:15'''
|''' 14:45-15:15'''
|'''[[Media:LUG2025-Amazon_FSx_Lustre_Open_Source-Day.pdf|Amazon FSx for Lustre and Open Source]]'''
|'''[[Media:LUG2025-Amazon_FSx_Lustre_Open_Source-Day.pdf|Amazon FSx for Lustre and Open Source]]''' ([https://www.youtube.com/watch?v=ZNIUdOioyMQ&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=11&pp=iAQB0gcJCbEJAYcqIYzv video])
:  Timothy Day, AWS
:  Timothy Day, AWS


|-
|-
|''' 15:30-16:00'''
|''' 15:30-16:00'''
|'''[[Media:LUG2025-Data_Loader_Caching_Hybrid_IO_Approach-Mishra.pdf|Impact of Data Loader Caching on Lustre: A Hybrid I/O Approach]]'''  
|'''[[Media:LUG2025-Data_Loader_Caching_Hybrid_IO_Approach-Mishra.pdf|Impact of Data Loader Caching on Lustre: A Hybrid I/O Approach]]''' ([https://www.youtube.com/watch?v=t1T9S2zo0j0&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=10&pp=iAQB video])
:  Rajeev Mishra, Hewlett Packard Enterprise
:  Rajeev Mishra, Hewlett Packard Enterprise
|-
|-
|''' 16:00-17:00'''
|''' 16:00-17:00'''
|'''[[Media:LUG2025-OpenSFS_Annual_Meeting-Larko.pdf|OpenSFS Annual Meeting]]'''
|'''[[Media:LUG2025-OpenSFS_Annual_Meeting-Larko.pdf|OpenSFS Annual Meeting]]''' ([https://www.youtube.com/watch?v=8cSJT2SAfQg&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=9&pp=iAQB video])
:  Megan Larko, OpenSFS Board
:  Megan Larko, OpenSFS Board


Line 67: Line 68:
|}
|}


=== Wednesday, May 8 – Day 2 ===
=== Wednesday, Apr 2 – Day 2 ===
{| border=1 cellpadding=0
{| border=1 cellpadding=0


|-
|-
|''' 09:00-10:00'''
|''' 09:00-10:00'''
|'''[[Media:LUG2025-Lustre_2.17_and_Beyond-Dilger.pdf|Lustre 2.17 and Beyond]]'''
|'''[[Media:LUG2025-Lustre_2.17_and_Beyond-Dilger.pdf|Lustre 2.17 and Beyond]]''' ([https://www.youtube.com/watch?v=qdgvGtxuxb8&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=8&pp=iAQB video])
:  Andreas Dilger, Whamcloud/DDN
:  Andreas Dilger, Whamcloud/DDN


|-
|-
|''' 10:30-11:00'''
|''' 10:30-11:00'''
|'''[[Media:LUG2025-TASSI-Bent.pdf|Rocks, Rabbits and Snakes, Oh my! - Lustre in LC]]'''
|'''[[Media:LUG2025-Rocks_Rabbits_Snakes_Lustre_in_LC-Harr.pdf|Rocks, Rabbits and Snakes, Oh my! - Lustre in LC]]''' ([https://www.youtube.com/watch?v=Evdo4GSstPg&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=7&pp=iAQB video])
:  Cameron Harr, Lawrence Livermore National Laboratory
:  Cameron Harr, Lawrence Livermore National Laboratory


|-
|-
|''' 11:00-11:30'''
|''' 11:00-11:30'''
|'''[[Media:LUG2025-Lustre_Multitenancy-Buisson.pdf|Lustre Multitenancy]]'''
|'''[[Media:LUG2025-Lustre_Multitenancy-Buisson.pdf|Lustre Multitenancy]]''' ([https://www.youtube.com/watch?v=IdAoED34n1g&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=6&pp=iAQB video])
:  Sebastien Buisson, Whamcloud/DDN
:  Sebastien Buisson, Whamcloud/DDN


|-
|-
|''' 11:30-12:00'''
|''' 11:30-12:00'''
|'''[[Media:LUG2025-Lustre_DNE3-Busting_the_Small_Files_Myth-Crusan.pdf|Hybrid IO Path Update]]'''
|'''[[Media:LUG2025-Lustre_DNE3-Busting_the_Small_Files_Myth-Crusan.pdf|Lustre DNE3 - Busting the Small Files Myth]]''' ([https://www.youtube.com/watch?v=Ty2NraEI3zI&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=5&pp=iAQB video])
:  Steve Crusan, Hudson River Trading
:  Steve Crusan, Hudson River Trading


|-
|-
|''' 13:00-13:15'''
|''' 13:00-13:15'''
|'''Sponsor Talk: [[Media:LUG2025-AWS_Sponsor_Talk.pdf|AWS Sponsor Talk]]'''
|'''Sponsor Talk: AWS Sponsor Talk'''  
:  AWS
:  AWS


|-
|-
|''' 13:15-13:45'''
|''' 13:15-13:45'''
|'''[[Media:LUG2025-AI_ML_Benchmarking_and_Lustre-Samar.pdf|All-flash Multinode High Availability for Lustre Disaggregated Implementations]]'''
|'''[[Media:LUG2025-All_Flash_Multinode_HA_Lustre_Disaggregated_Implementations-Landau-Villa.pdf|All-flash Multinode High Availability for Lustre Disaggregated Implementations]]''' ([https://www.youtube.com/watch?v=vde4aagsnU8&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=4&pp=iAQB0gcJCbEJAYcqIYzv video])
:  Daniel Landau, Xinnor; Davide Villa, Xinnor
:  Daniel Landau, Xinnor; Davide Villa, Xinnor


|-
|-
|''' 13:45-14:15'''
|''' 13:45-14:15'''
|'''[[Media:LUG2025-NCSA_Lustre_Site_Update-Maloney.pdf|NCSA Lustre Site Update]]'''
|'''[[Media:LUG2025-NCSA_Lustre_Site_Update-Maloney.pdf|NCSA Lustre Site Update]]''' ([https://www.youtube.com/watch?v=3jqjMRPxlnY&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=3&pp=iAQB video])
:  J.D. Maloney, National Center for Supercomputing Applications
:  J.D. Maloney, National Center for Supercomputing Applications


|-
|-
|''' 14:15-14:45'''
|''' 14:15-14:45'''
|'''[[Media:LUG2025-Lustre_Timeout_Hierarchy-Horn.pdf|Lustre Timeout Hierarchy]]'''
|'''[[Media:LUG2025-Lustre_Timeout_Hierarchy-Horn.pdf|Lustre Timeout Hierarchy]]''' ([https://www.youtube.com/watch?v=VbDAeuo5534&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=2&pp=iAQB video])
:  Chris Horn, Hewlett Packard Enterprise
:  Chris Horn, Hewlett Packard Enterprise


|-
|-
|''' 15:15-15:45'''
|''' 15:15-15:45'''
|'''[[Media:LUG2025-Robinhood_4_Policy_Engine_Toolbox-Courrier.pdf|Robinhood 4: The Policy Engine Toolbox]]'''
|'''[[Media:LUG2025-Robinhood_4_Policy_Engine_Toolbox-Courrier.pdf|Robinhood 4: The Policy Engine Toolbox]]''' ([https://www.youtube.com/watch?v=uXdEmsM4Whc&list=PLA5dHg1_l3V-ceplO8QAJVRIsR0Fh85Cu&index=1&pp=iAQB video])
:  Guillaume Courrier, CEA
:  Guillaume Courrier, CEA



Latest revision as of 05:46, 1 June 2025

LUG 2025 was held at Stanford Research Computing in Stanford, California on April 1-2 2025

LUG 2025 Agenda

Tuesday, Apr 1 – Day 1

09:15-09:30 Welcoming Remarks
Stephane Thiell, Stanford; Megan Larko, OpenSFS
09:30-10:00 Unraveling the Universe with High Performance Computing (video)
Marcelo Alvarez, Kavli Institute for Particle Astrophysics and Cosmology
10:00-10:30 Lustre Community Update (video)
Peter Jones, Whamcloud
11:00-11:30 Streamlining AI Checkpoints: Automatic Data Migration to Object Store (video)
Aurélien Degrémont, NVIDIA
11:30-12:00 Lustre Upstreaming Efforts (video)
James Simmons, Oak Ridge National Laboratory
13:00-13:15 Sponsor Talk: EXAScaler Feature Overview (video)
Ivan Poddubnyy, DDN; Matt Rásó-Barnett, DDN
13:15-13:45 ALCF Site Update (video)
Alex Kulyavtsev, Argonne National Laboratory
13:45-14:15 Sandia Labs Lustre Filesystem Customer IO DOS Mitigation Using NRS Token Bucket Filters (video)
Michael Aguilar, Sandia National Laboratories; Jason Peter, Sandia National Laboratories
14:45-15:15 Amazon FSx for Lustre and Open Source (video)
Timothy Day, AWS
15:30-16:00 Impact of Data Loader Caching on Lustre: A Hybrid I/O Approach (video)
Rajeev Mishra, Hewlett Packard Enterprise
16:00-17:00 OpenSFS Annual Meeting (video)
Megan Larko, OpenSFS Board
17:30-21:30 Networking and Social Event
Barebottle Brewing Co.

Wednesday, Apr 2 – Day 2

09:00-10:00 Lustre 2.17 and Beyond (video)
Andreas Dilger, Whamcloud/DDN
10:30-11:00 Rocks, Rabbits and Snakes, Oh my! - Lustre in LC (video)
Cameron Harr, Lawrence Livermore National Laboratory
11:00-11:30 Lustre Multitenancy (video)
Sebastien Buisson, Whamcloud/DDN
11:30-12:00 Lustre DNE3 - Busting the Small Files Myth (video)
Steve Crusan, Hudson River Trading
13:00-13:15 Sponsor Talk: AWS Sponsor Talk
AWS
13:15-13:45 All-flash Multinode High Availability for Lustre Disaggregated Implementations (video)
Daniel Landau, Xinnor; Davide Villa, Xinnor
13:45-14:15 NCSA Lustre Site Update (video)
J.D. Maloney, National Center for Supercomputing Applications
14:15-14:45 Lustre Timeout Hierarchy (video)
Chris Horn, Hewlett Packard Enterprise
15:15-15:45 Robinhood 4: The Policy Engine Toolbox (video)
Guillaume Courrier, CEA
15:45-16:45 The Future of Lustre Panel Q&A
Julie Bernauer, Director of Data Center Systems Engineering at NVIDIA
Andreas Dilger, Lustre Principal Architect at Whamcloud/DDN
Cameron Harr, Lustre Operations Lead and I/O Strategist at Lawrence Livermore National Laboratory
J.D. Maloney, Lead HPC Storage Engineer at the National Center for Supercomputing Applications
16:45-17:00 Closing Remarks
Megan Larko

Thursday, Apr 3 – Day 3

Developer Day