BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260522T150123Z
LOCATION:D174
DTSTART;TZID=America/Chicago:20181116T094000
DTEND;TZID=America/Chicago:20181116T100000
UID:submissions.supercomputing.org_SC18_sess146_ws_ftxs112@linklings.com
SUMMARY:A Comprehensive Informative Metric for Analyzing HPC System Status
  Using the LogSCAN Platform
DESCRIPTION:Yawei Hui, Byung Hoon Park, and Christian Engelmann (Oak Ridge
  National Laboratory)\n\nLog processing by Spark and Cassandra-based ANaly
 tics (LogSCAN) is a newly developed analytical platform that provides flex
 ible and scalable data gathering, transformation and computation. One majo
 r challenge is to effectively summarize the status of a complex computer s
 ystem, such as the Titan supercomputer at the Oak Ridge Leadership Computi
 ng Facility (OLCF). Although there is plenty of operational and maintenanc
 e information collected and stored in real time, which may yield insights 
 about short- and long-term system status, it is difficult to present this 
 information in a comprehensive form. In this work, we present system infor
 mation entropy (SIE), a newly developed metric that leverages the powers o
 f traditional machine learning techniques and information theory. By compr
 essing the multi-variant multi-dimensional event information recorded duri
 ng the operation of the targeted system into a single time series of SIE, 
 we demonstrate that the historical system status can be sensitively repres
 ented concisely and comprehensively. Given a sharp indicator as SIE, we ar
 gue that follow-up analytics based on SIE will reveal in-depth knowledge a
 bout system status using other sophisticated approaches, such as pattern r
 ecognition in the temporal domain or causality analysis incorporating extr
 a independent metrics of the system.\n\nTag: Resiliency, Scientific Comput
 ing\n\nRegistration Category: Workshop Reg Pass\n\n
END:VEVENT
END:VCALENDAR