BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20141118T170000Z
DTEND:20141118T173000Z
LOCATION:291
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: Data stored by many HPC organizations is increasing exponentially.  And while today’s high capacity scale-out file and object storage systems can easily accommodate the growing volume of data, their limited, bare-bones system metadata can leave high-value data lost over time – stranding it and losing its value forever. Without application-specific metadata, HPC organizations will amass a growingly unmanageable “Data Junkyard” of inaccessible, ultimately useless data, at very high cost.  This presentation will explore how the Big Data challenges faced by Big Science projects like CERN’s Large Hadron Collider, and the metadata strategies those projects pioneered, could be exploited to support data intensive HPC workflows by using the metadata management facilities of the Nirvana Metadata Centric Intelligent Storage system. Nirvana was developed by the General Atomics Magnetic Fusion Energy and Advanced Concepts group in San Diego, California from a joint effort with the San Diego Supercomputing Center’s Storage Resource Broker (SRB).
SUMMARY:Eliminating Data Junkyards with Metadata
PRIORITY:3
END:VEVENT
END:VCALENDAR