Lustre User Group 2016

LUG 2016 was held in Portland, Oregon, from April 5th-7th, 2016. Additionally, Lustre Developer Day 2016 was held on April 4th, 2016.

LUG 2016 Agenda

Day 1 – Tuesday, April 5

• Welcome Remarks - Stephen Simms, LUG Program Chair
• Lustre 101: A Quick Overview - Stephen Simms, LUG Program Chair
• Community Release Update - Peter Jones, LWG Co-Chair
• Lustre 2.9 and Beyond - Andreas Dilger, Intel
• Improved Versioning, Building, Packaging, and Distribution of Lustre - Christopher Morrone, Lawrence Livermore National Laboratory, Giuseppe Di Natale, Lawrence Livermore National Laboratory
• LMT: It’s Only A Flesh Wound - Olaf Faaland, Lawrence Livermore National Laboratory
• Enhancing Lustre Security with the Whitelist/Blacklist Patch - Josh Judd, Warp Mechanics
• Lustre Security Infrastructure Today and Tomorrow - John Hammond, Intel
• Security Isolation for Lustre - Sebastien Buisson, DDN
• Lustre Update from Seagate - Peter Bojanic, Seagate
• Large Streaming IO for Lustre on ZFS - Jinshan Xiong, Intel
• Lustre ZFS Snaphots - Fan Yong, Intel
• Vectorized ZFS RAIDZ Implementation - Gvozden Nešković, Frankfurt Institute for Advanced Studies
• Status of the Upstream Client - James Simmons, Oak Ridge National Laboratory
• Removing Technical Debt - Ben Evans, Cray
• Intel(r) Solutions for Lustre* Software: Update - Jessica Popp, Intel
• Lustre.Org – A Community Resource - Ken Rawlings, Lustre.org Working Group OpenSFS Liaison
• OpenSFS and EOFS Update - Charlie Carroll, Cray, OpenSFS Board Member, Hugo Falter, EOFS Board Member

Day 2 – Wednesday, April 6

• Lustre in a Condo Computing Environment - John White, Lawrence Berkeley National Laboratory
• Lustre Deployed Three Different Ways to Meet Researchers’ Needs - Scott Yockel, Harvard University
• Tiering Storage (around Lustre) for Big Data and Supercomputing: Cray’s Portfolio of Offerings - Jason Goodman, Cray Inc.
• The 1 Million IOPS Lustre File System at TU Dresden - Michael Kluge, Technische Universität Dresden Johann Peyrard, Atos
• Building Lustre Solutions with NVM Devices Today - James Coomer, DataDirect Networks
• Lustre Deployment using Intel OmniPath Interconnect - Brian Johanson, Pittsburgh Supercomputing Center J. Ray Scott, Pittsburgh Supercomputing Center
• InfiniBand At A Distance - Steve Woods, Cray, Dave McMillen, Cray
• Multi-rail LNet for Lustre - Amir Shehata, Intel Olaf Weber, SGI
• A New Approach to Lustre – Join HPE for a Look at the Future of Optimized Lustre Environments - Craig Belusar, Hewlett Packard Enterprise
• An Architecture for Docker on Lustre - Blake Caldwell, Oak Ridge National Laboratory
• Scaling Apache Spark on Lustre - Nicholas Chaimov, University of Oregon
• Scaling LDISKFS for the Future - Artem Blagodarenko, Seagate
• DL-SNAP: A Directory Level SNAPSHOT Facility on Lustre - Shinji Sumimoto, Fujitsu
• Board Q&A Panel

Day 3 – Thursday, April 7

• Developing an Open Source Object Storage Copytool for HSM on Lustre - Frederick Lefebvre, Calcul Québec Simon Guilbault, Calcul Québec
• Lustre Data Mover: Because File Systems are Rooted Trees and rsync Must Die - Rick Wagner, San Diego Supercomputer Center
• Project Quota for Lustre - Li Xi, DDN
• ASCAR: Increasing Performance Through Automated Contention Management - Yan Li, University of California, Santa Cruz
• Evaluating Progressive File Layouts for Lustre - Richard Mohr, University of Tennessee, Knoxville
• Closing - Stephen Simms, LUG Program Chair