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:20210402T160556Z
LOCATION:Track 5
DTSTART;TZID=America/New_York:20201112T110500
DTEND;TZID=America/New_York:20201112T113200
UID:submissions.supercomputing.org_SC20_sess211_ws_pdsw101@linklings.com
SUMMARY:Keeping It Real: Why HPC Data Services Don't Achieve I/O Microbenc
 hmark Performance
DESCRIPTION:Workshop\n\nKeeping It Real: Why HPC Data Services Don't Achie
 ve I/O Microbenchmark Performance\n\nCarns, Harms, Settlemyer, Atkinson, R
 oss\n\nHPC storage software developers rely on benchmarks as reference poi
 nts for performance evaluation. Low-level synthetic microbenchmarks are pa
 rticularly valuable for isolating performance bottlenecks in complex syste
 ms and identifying optimization opportunities.\n\nThe use of low-level mic
 robenchmarks also entails risk, however, especially if the benchmark behav
 ior does not reflect the nuances of production data services or applicatio
 ns. In those cases, microbenchmark measurements can lead to unrealistic ex
 pectations or misdiagnosis of performance problems.  Neither benchmark cre
 ators nor software developers are necessarily at fault in this scenario, h
 owever.  The underlying problem is more often a subtle disconnect between 
 the objective of the benchmark and the objective of the developer.\n\nIn t
 his paper, we investigate examples of discrepancies between microbenchmark
  behavior and software developer expectations. Our goal is to draw attenti
 on to these pitfalls and initiate a discussion within the community about 
 how to improve the state of the practice in performance engineering for HP
 C data services.\n\nTag: Big Data, Data Analytics, Compression, and Manage
 ment, Data Movement, File Systems and I/O, Storage\n\nRegistration Categor
 y: Workshop Reg Pass
END:VEVENT
END:VCALENDAR