Lustre User Group 2013
Jump to navigation
Jump to search
Sun, sand and Lustre! To celebrate one of the biggest years for Lustre – multiple new Lustre installations and now clocking at over 1TB/sec – LUG ’13 successfully took place in San Diego at the Omni Hotel in the Downtown Gaslamp Quarter. Sessions were held from April 16-18.
LUG 2013 Agenda
Tuesday, April 16, 2013 - LUG Day 1 - Omni Hotel
7:00 - 8:00 | Breakfast |
8:00 - 8:10 | Welcome Remarks |
8:10 - 8:45 | OpenSFS Update
|
8:45 - 9:20 | EOFS Update
|
9:20 - 9:30 | Xyratex Update
|
9:30 - 10:00 | Lustre Releases
|
10:00 - 10:20 | Break |
10:20 - 10:40 | Layout Lock
|
10:40 - 11:00 | Distributed Namespace (DNE)
|
11:00 - 11:20 | Software vs. Hardware RAID and Implications for the Future
|
11:20 - 11:40 | Sequoia and the ZFS OSD
|
11:40 - 12:00 | ZFS and Lustre go to Hollywood
|
12:00 - 1:00 | Lunch |
1:00 - 1:20 | Lustre Contribution Model
|
1:20 - 1:40 | The State of the Lustre File System and The Lustre Development Ecosystem: A 2013 Report Card
|
1:40 - 2:00 | Lustre Acquisition and Its Future
|
2:00 - 2:40 | OpenSFS CDWG Meeting
|
2:40 - 3:00 | The Lustre Community in Europe
|
3:00 - 3:30 | Break |
3:30 - 3:50 | Lustre Tuning Parameters
|
3:50 - 4:10 | Hadoop-Lustre
|
4:10 - 4:30 | An Intro to Ceph for HPC
|
4:30 - 4:50 | Lustre on Amazon Web Services
|
4:50 - 6:00 | Vendor Presentations
|
6:00 | Adjourn |
Wednesday, April 17, 2013 - LUG Day 2 - Omni Hotel
7:00 - 8:00 | Breakfast |
8:00 - 8:30 | Lustre 2.5 and Beyond
|
8:30 - 8:50 | DAOS Changes to Lustre
|
8:50 - 9:10 | Online LFSCK
|
9:10 - 10:00 | OpenSFS TWG Meeting
|
10:00 - 10:20 | Break |
10:20 - 10:40 | User-Defined Transport Protocols for Lustre
|
10:40 - 11:00 | Exploring Multiple Interface Lustre Performance from a Single Client
|
11:00 - 11:20 | Active-active LNET Bonding Using Multiple LNETs and Infiniband partitions
|
11:20 - 11:40 | Getting the Most Out of Lustre with the TCP LND
|
11:40 - 12:00 | Wireshark
|
12:00 - 1:00 | Lunch |
1:00 - 2:00 | Lustre in Commercial Space Panel |
2:00 - 2:20 | Lustre Static Code Analysis with Coverity
|
2:20 - 2:40 | Lustre Test and Validation Toolset
|
2:40 - 3:00 | Lustre Manual
|
3:00 - 3:30 | Break |
3:30 - 3:50 | Overcoming Gemini LNET Performance Hurdles
|
3:50 - 4:10 | Performance & Functionality Testbed for Clustered Filesystems: Lustre and some of its friends
|
4:10 - 4:30 | A New Metric for File System Load
|
4:30 - 4:50 | HPCS Scenarios Update
|
4:50 - 5:50 | OpenSFS BWG Meeting
|
5:50 | Adjourn |
Thursday, April 18, 2013 - LUG Day 3 - Omni Hotel
7:00 - 8:00 | Breakfast |
8:00 - 8:20 | The next-generation 1 TB/s Spider file system at OLCF
|
8:20 - 8:40 | High Availability in Lustre - Enterprise RAS
|
8:40 - 9:00 | Managing and Monitoring a Scalable Lustre Infrastructure
|
9:00 - 9:20 | Robinhood Policy Engine
|
9:20 - 9:40 | Sequoia Data Migration Experiences
|
9:40 - 10:00 | Lustre HSM & Cloud: An Open Discussion on the Future of Lustre HSM,
Storage Tiering, File/Block/Object Backing Stores & Lustre File Geo-Distribution
|
10:00 - 10:30 | Break |
10:30 - 11:00 | OpenSFS General Assembly
|
11:00 - 11:20 | Fujitsu Contributions to Lustre
|
11:20 - 11:40 | Using Changelogs for Efficient Search and Content Discovery
|
11:40 - 12:00 | A Need for a DNS-like feature for LNet NIDs
|
12:00 - 1:00 | Lunch |
1:00 - 1:30 | Next Generation Storage Architectures for Exascale
|
1:30 - 2:00 | EIOW - a Framework for Exa-Scale I/O
|
2:00 - 2:30 | Performance Evaluation of FEFS on K Computer and Fujitsu’s Roadmap toward Lustre 2.x
|
2:30 - 3:00 | Lustre - Fast Forward to Exascale
|
3:00 - 3:30 | Break |
3:30 - 4:30 | Lustre and Beyond Panel |
4:30 - 4:50 | Lustre at FNAL
|
4:50 - 5:10 | Lustre at BlueWaters
|
5:10 - 5:30 | SDSC’s Data Oasis: Balanced Performance and Cost-Effective Lustre File Systems
|
5:30 - 5:50 | The Madness of Project George
|
5:50 | End of LUG 2013 |