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:20260422T000711Z
LOCATION:704-706
DTSTART;TZID=America/Denver:20231113T113900
DTEND;TZID=America/Denver:20231113T115700
UID:submissions.supercomputing.org_SC23_sess450_ws_worksp122@linklings.com
SUMMARY:Distributed Data Locality-Aware Job Allocation
DESCRIPTION:Ana Markovic, Dimitris Kolovos, and Leandro Soares Indrusiak (
 University of York, England)\n\nScheduling tasks close to their associated
  data is crucial in distributed systems to minimize network traffic and la
 tency. Some Big Data frameworks like Apache Spark employ locality function
 s and job allocation algorithms to minimize network traffic and execution 
 times. However, these frameworks rely on centralized mechanisms, where the
  master node determines data locality by allocating tasks to available wor
 kers with minimal data transfer time, ignoring variances in worker configu
 rations and availability. To address these limitations, we propose a decen
 tralized approach to locality-driven scheduling that grants workers autono
 my in the job allocation process while factoring in workers' configuration
 s, such as network and CPU speed differences. Our approach is developed an
 d evaluated on Crossflow, a distributed stream processing platform with da
 ta-aware independent worker nodes. Preliminary evaluation experiments indi
 cate that our approach can yield up to 3.57x faster execution times when c
 ompared to the baseline centralized approach where the master controls dat
 a locality.\n\nTag: Data Analysis, Visualization, and Storage, Large Scale
  Systems, Programming Frameworks and System Software, Reproducibility, Res
 ource Management, Runtime Systems\n\nRegistration Category: Workshop Reg P
 ass\n\nSession Chairs: Silvina Caino-Lores (National Institute for Researc
 h in Digital Science and Technology (Inria)) and Anirban Mandal (Renaissan
 ce Computing Institute (RENCI), University of North Carolina at Chapel Hil
 l)\n\n
END:VEVENT
END:VCALENDAR