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:20210402T160546Z
LOCATION:Track 9
DTSTART;TZID=America/New_York:20201113T152500
DTEND;TZID=America/New_York:20201113T155500
UID:submissions.supercomputing.org_SC20_sess231_ws_mlcs101@linklings.com
SUMMARY:A Year of Automated Anomaly Detection in a Datacenter
DESCRIPTION:Workshop\n\nA Year of Automated Anomaly Detection in a Datacen
 ter\n\nAhmed, Porter, Abdelmutalab, Ricci\n\nAnomaly detection based on ma
 chine learning can be a powerful tool for understanding the behavior of la
 rge, complex computer systems in the wild. The set of anomalies seen, howe
 ver, can change over time: as the system evolves, is put to different uses
  and encounters different workloads, both its typical behavior and the ano
 malies that it encounters can change as well. This naturally raises two qu
 estions: how effective is automated anomaly detection in this setting, and
  how much does anomalous behavior change over time? \n\nIn this paper, we 
 examine these questions for a dataset taken from a system that manages the
  lifecycle of servers in datacenters. We look at logs from one year of ope
 ration of a datacenter of about 500 servers. Applying state-of-the art tec
 hniques for finding anomalous events, we find that there are a core set of
  anomaly patterns that persist over the entire period studied, but that to
  track the evolution of the system, we must re-train the detector periodic
 ally. Working with the administrators of this system, we find that, despit
 e these changes in patterns, they still contain actionable insights.\n\nRe
 gistration Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR