BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:1.0
BEGIN:VEVENT
DTSTART:20111117T013000Z
DTEND:20111117T030000Z
LOCATION:TCC 305
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: Achieving efficient data movement over high-speed networks remains an ongoing challenge. Factors that affect end-to-end performance can exist in host systems, within local-area networks, across wide-area paths, and at storage systems. The problem space consists of a highly complex set of layers and domains, including application software, middleware, OS software, system hardware, network software/hardware, file systems and storage. What is worse, the nature of these performance problems is very different, compared to those we are all trained to watch for and diagnose. We need tools that can detect, localize and diagnose transient bottlenecks in complex end-to-end distributed systems. =0A=0AThis BoF will focus on the following questions:=0A=0A(1)	What are the performance problems that take place in high-end distributed systems? =0A=0A(2)	How can such performance problems be detected, diagnosed and localized? =0A =0A(3)	How can existing network performance monitoring and diagnosis tools/services scale to high-speed networks?
SUMMARY:End-to-End Performance Problems in High-speed Networks
PRIORITY:3
END:VEVENT
END:VCALENDAR