11.12.2012 Views

NASA Scientific and Technical Aerospace Reports

NASA Scientific and Technical Aerospace Reports

NASA Scientific and Technical Aerospace Reports

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

version compatibly issues surfacing once again. Maturity will take some time.<br />

Author<br />

Data Storage; Stockpiling; Data Management; Consolidation<br />

20040121058 Paderborn Univ., Germany<br />

V:Drive: Costs <strong>and</strong> Benefits of an Out-of-B<strong>and</strong> Storage Virtualization System<br />

Brinkmann, Andre; Heidebuer, Michael; MeyeraufderHeide, Friedhelm; Ruekert, Ulrich; Salzwedel, Kay; Vodisek, Mario;<br />

<strong>NASA</strong>/IEEE MSST 2004 Twelfth <strong>NASA</strong> Goddard Conference on Mass Storage Systems <strong>and</strong> Technologies in cooperation<br />

with the Twenty-First IEEE Conference on Mass Storage Systems <strong>and</strong> Technologies; April 2004, pp. 153-157; In English; See<br />

also 20040121020<br />

Contract(s)/Grant(s): DFG-IST-1999-14186; No Copyright; Avail: CASI; A01, Hardcopy<br />

The advances in network technology <strong>and</strong> the growth of the Internet together with upcoming new applications like<br />

peer-to-peer (P2P) networks have led to an exponential growth of the stored data volume. The key to manage this data<br />

explosion seems to be the consolidation of storage systems inside storage area networks (SANs) <strong>and</strong> the use of a storage<br />

virtualization solution that is able to abstract from the underlying physical storage system. In this paper we present the first<br />

measurements on an out-of-b<strong>and</strong> storage virtualization system <strong>and</strong> investigate its performance <strong>and</strong> scalability compared to a<br />

plain SAN. We show in general that a carefully designed out-of-b<strong>and</strong> solution has only a very minor impact on the CPU usage<br />

in the connected servers <strong>and</strong> that the metadata management can be efficiently h<strong>and</strong>led. Furthermore we show that the use of<br />

an adaptive data placement scheme in our virtualization solution V:Drive can significantly enhance the throughput of the<br />

storage systems, especially in environments with r<strong>and</strong>om access schemes.<br />

Author<br />

Data Storage; Computer Storage Devices<br />

20040121059 California Univ., Lawrence Berkeley National Lab., Berkeley, CA, USA<br />

GUPFS: The Global Unified Parallel File System Project at NERSC<br />

Butler, Greg; Lee, Rei; Welcome, Mike; <strong>NASA</strong>/IEEE MSST 2004 Twelfth <strong>NASA</strong> Goddard Conference on Mass Storage<br />

Systems <strong>and</strong> Technologies in cooperation with the Twenty-First IEEE Conference on Mass Storage Systems <strong>and</strong> Technologies;<br />

April 2004, pp. 361-371; In English; See also 20040121020<br />

Contract(s)/Grant(s): DE-AC03-76SF-00098; No Copyright; Avail: CASI; A03, Hardcopy<br />

The Global Unified Parallel File System (GUPFS) project is a five -year project to provide a scalable, high -performance,<br />

high -b<strong>and</strong>width, shared file system for the National Energy Research <strong>Scientific</strong> Computing Center (NERS C). This paper<br />

presents the GUPFS testbed configuration, our benchmarking methodology, <strong>and</strong> some preliminary results.<br />

Author<br />

Information Transfer; Computer Storage Devices; Computer Networks<br />

20040121060 California Univ., Santa Cruz, CA, USA<br />

Duplicate Data Elimination in a SAN File System<br />

Hong, Bo; Long, Darrell D. E.; Plantenberg, Demyn; Sivan-Zimet, Miriam; <strong>NASA</strong>/IEEE MSST 2004 Twelfth <strong>NASA</strong> Goddard<br />

Conference on Mass Storage Systems <strong>and</strong> Technologies in cooperation with the Twenty-First IEEE Conference on Mass<br />

Storage Systems <strong>and</strong> Technologies; April 2004, pp. 301-314; In English; See also 20040121020; No Copyright; Avail: CASI;<br />

A03, Hardcopy<br />

Duplicate Data Elimination (DDE) is our method for identifying <strong>and</strong> coalescing identical data blocks in Storage Tank, a<br />

SAN file system. On-line file systems pose a unique set of performance <strong>and</strong> implementation challenges for this feature.<br />

Existing techniques, which are used to improve both storage <strong>and</strong> network utilization, do not satisfy these constraints. Our<br />

design employs a combination of content hashing, copy-on-write, <strong>and</strong> lazy updates to achieve its functional <strong>and</strong> performance<br />

goals. DDE executes primarily as a background process. The design also builds on Storage Tank’s FlashCopy function to ease<br />

implementation. We include an analysis of selected real-world data sets that is aimed at demonstrating the space-saving<br />

potential of coalescing duplicate data. Our results show that DDE can reduce storage consumption by up to 80% in some<br />

application environments. The analysis explores several additional features, such as the impact of varying file block size <strong>and</strong><br />

the contribution of whole file duplication to the net savings.<br />

Author<br />

Deletion; Data Processing; Data Storage<br />

237

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!