BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260522T150119Z
LOCATION:C146
DTSTART;TZID=America/Chicago:20181113T103000
DTEND;TZID=America/Chicago:20181113T110000
UID:submissions.supercomputing.org_SC18_sess179_pap165@linklings.com
SUMMARY:SP-Cache: Load-Balanced, Redundancy-Free Cluster Caching with Sele
 ctive Partition
DESCRIPTION:Yinghao Yu, Renfei Huang, Wei Wang, Jun Zhang, and Khaled Ben 
 Letaief (Hong Kong University of Science and Technology)\n\nData-intensive
  clusters increasingly employ in-memory solutions to improve I/O performan
 ce. However, the routinely observed file popularity skew and load imbalanc
 e create hotspots, which significantly degrades the benefits of in-memory 
 solutions. Common approaches to tame load imbalance include copying multip
 le replicas of hot files and creating parity chunks using storage codes. Y
 et, these techniques either suffer from high memory redundancy or incur no
 n-trivial encoding/decoding overhead. In this paper, we propose a differen
 t approach to achieve load balancing without memory redundancy or encoding
 /decoding overhead. Our solution, termed SP-Cache, selectively partitions 
 files based on their popularity and evenly caches those partitions across 
 the cluster. We develop an efficient algorithm to determine the optimal nu
 mber of partitions for hot files—too few partitions are incapable of mitig
 ating hotspots, while too many are susceptible to stragglers. EC2 deployme
 nt and trace-driven simulations show that, compared with existing solution
 s, SP-Cache reduces the read latencies by up to 40%.\n\nTag: Clouds and Di
 stributed Computing, File Systems, I/O, Storage\n\nRegistration Category: 
 Tech Program Reg Pass\n\nSession Chair: Dennis Gannon (Indiana University,
  The eScience Cloud)\n\n
END:VEVENT
END:VCALENDAR