BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240116T191656Z
LOCATION:704-706
DTSTART;TZID=America/Denver:20231113T093700
DTEND;TZID=America/Denver:20231113T095500
UID:submissions.supercomputing.org_SC23_sess450_ws_worksp117@linklings.com
SUMMARY:A Data Science Pipeline Synchronization Method for Edge-Fog-Cloud 
 Continuum
DESCRIPTION:Workshop\n\nDante D. Sanchez-Gallegos (University Carlos III o
 f Madrid, Spain); J. L. Gonzalez-Compean (Cinvestav Tamaulipas); Jesus Car
 retero (University Carlos III of Madrid, Spain); and Heidy Marin-Castro (C
 átedras CONACYT - Universidad Autónoma de Tamaulipas)\n\nThis paper presen
 ts an adaptive continuum synchronization method for data science pipelines
  deployed on edge-fog-cloud infrastructures. In a diagnostic phase, a mode
 l, based on the Bernoulli principle, is used as an analogy to create a glo
 bal representation of bottlenecks in a pipeline. In a supervision phase, a
  watchman/sentinel cooperative system monitors and captures the throughput
  of the pipeline stages to create a bottleneck-stage scheme. In a rectific
 ation phase, this system produces replicas of stages identified as bottlen
 ecks to mitigate the workload congestion using implicit parallelism and lo
 ad balancing algorithms. This method is automatically and transparently in
 voked to produce in runtime a steady continuum dataflow. To test our propo
 sal, we conducted a case study about the processing of medical and satelli
 te data on fog-cloud infrastructures. The evaluation revealed that this me
 thod creates, without characterizing workloads nor knowing infrastructure 
 details, continuum dataflows, which yield a competitive performance with s
 olutions in the state-of-the-art.\n\nTag: Data Analysis, Visualization, an
 d Storage, Large Scale Systems, Programming Frameworks and System Software
 , Reproducibility, Resource Management, Runtime Systems\n\nRegistration Ca
 tegory: Workshop Reg Pass\n\nSession Chairs: Silvina Caino-Lores (French I
 nstitute for Research in Computer Science and Automation (INRIA)) and Anir
 ban Mandal (Renaissance Computing Institute (RENCI), University of North C
 arolina)
END:VEVENT
END:VCALENDAR