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:405-406-407
DTSTART;TZID=America/Denver:20231115T153000
DTEND;TZID=America/Denver:20231115T160000
UID:submissions.supercomputing.org_SC23_sess183_pap155@linklings.com
SUMMARY:Rethinking Deployment for Serverless Functions: A Performance-Firs
 t Perspective
DESCRIPTION:Yiming Li, Laiping Zhao, Yanan Yang, and Wenyu Qu (Tianjin Uni
 versity)\n\nServerless computing commonly adopts strong isolation mechanis
 ms for deploying functions, which may bring significant performance overhe
 ad because each function needs to run in a completely new environment, i.e
 ., the “one-to-one” model. To accelerate the function computation, prior w
 ork has proposed using sandbox sharing to reduce the overhead, i.e., the “
 many-to-one” model. Nonetheless, either process-based true-parallelism or 
 thread-based pseudo-parallelism prevents its adaptation for latency-sensit
 ive web services. \n\nTo achieve optimal performance and resource efficien
 cy for serverless workflow, we argue an “m-to-n” deployment model that man
 ipulates multiple granularities of computing abstractions (e.g., processes
 , threads), and sandboxes to amortize overhead. We propose wrap, a new dep
 loyment abstraction that balances the tradeoffs between interaction overhe
 ad, startup overhead and function execution. We further design Chiron, a w
 rap-based deployment manager that can automatically perform the orchestrat
 ion of multiple computing abstractions based on performance prioritization
 . Our comprehensive evaluation indicates Chiron outperforms state-of-the-a
 rt systems by 1.3x-21.8x on system throughput.\n\nTag: Cloud Computing, Di
 stributed Computing, Energy Efficiency, Performance Measurement, Modeling,
  and Tools\n\nRegistration Category: Tech Program Reg Pass\n\nReproducibil
 ity Badges: Artifact Available, Artifact Functional, Results Reproduced\n\
 nSession Chair: Radu Prodan (University of Innsbruck, Austria)\n\n
END:VEVENT
END:VCALENDAR