Hardware Reference
In-Depth Information
Chapter 5
Livermore Computing Center
Richard Hedges and Blaise Barney
Livermore Computing Center, Lawrence Livermore National Laboratory
5.1
Introduction ::::::::::::::::::::::::::::::::::::::::::::::::::::::
51
5.2
The Lustre
R
Parallel File System: Early Developments ::::::::
53
5.3
Sequoia, Lustre
R
2.0, and ZFS ::::::::::::::::::::::::::::::::::
54
5.4
IBM Blue Gene Systems ::::::::::::::::::::::::::::::::::::::::
55
5.5
Sequoia File System Hardware :::::::::::::::::::::::::::::::::::
57
5.6
Experience with ZFS-Based Lustre
R
and Sequoia in Production
59
5.7
Sequoia I/O in Practice ::::::::::::::::::::::::::::::::::::::::::
60
5.7.1
General Remarks ::::::::::::::::::::::::::::::::::::::::
60
5.7.2
Recommendations to Application Developers :::::::::::
61
5.7.3
SILO: LLNL's I/O Library ::::::::::::::::::::::::::::::
62
5.7.4
Scalable Checkpoint/Restart ::::::::::::::::::::::::::::
62
5.8
Conclusion ::::::::::::::::::::::::::::::::::::::::::::::::::::::::
63
Bibliography ::::::::::::::::::::::::::::::::::::::::::::::::::::::
63
This contribution is a snapshot of the IBM Blue Gene/Q system, which is
presently the largest deployed system in the Livermore Computing Center
(LC) at the Lawrence Livermore National Laboratory (LLNL). The primary
focus of this chapter is to elaborate on the details of the file system component
of this system.
5.1 Introduction
LLNL is one of the research and development facilities of the Department
of Energy (DOE), founded in 1952, and is presently operated by Lawrence Liv-
ermore National Security. LLNL's general mission is to support the national
interest in science through the application of wide-ranging and advanced sci-
ence, engineering, and technology. One enduring critical mission is to ensure
the safety, security, and reliability of the nation's nuclear weapons.
51
