BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210402T160559Z
LOCATION:Track 7
DTSTART;TZID=America/New_York:20201112T110000
DTEND;TZID=America/New_York:20201112T112500
UID:submissions.supercomputing.org_SC20_sess213_ws_isav107@linklings.com
SUMMARY:Chimbuko: A Workflow-Level Scalable Performance Trace Analysis Too
 l
DESCRIPTION:Workshop\n\nChimbuko: A Workflow-Level Scalable Performance Tr
 ace Analysis Tool\n\nKelly, Ha, Huck, Van Dam, Pouchard...\n\nDue to the s
 heer volume of data, it is typically impractical to analyze the detailed p
 erformance of an HPC application running at-scale. While conventional smal
 l-scale benchmarking and scaling studies are often sufficient for simple a
 pplications, many modern workflow-based applications couple multiple eleme
 nts with competing resource demands and complex inter-communication patter
 ns for which performance cannot easily be studied in isolation and at smal
 l scale. This work discusses Chimbuko, a performance analysis framework th
 at provides real-time, in situ anomaly detection. By focusing specifically
  on performance anomalies and their origin (a.k.a provenance), data volume
 s are dramatically reduced without losing necessary details. To the best o
 f our knowledge, Chimbuko is the first online, distributed and scalable wo
 rkflow-level performance trace analysis framework. We demonstrate the tool
 's usefulness on Oak Ridge National Laboratory's Summit system.\n\nRegistr
 ation Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR