BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Scavenger: A Black-Box Batch Workload Resource Manager for Improvi
 ng Utilization in Cloud Environments - Seyyed Ahmad Javadi
DTSTART:20191119T131500Z
DTEND:20191119T134500Z
UID:TALK135238@talks.cam.ac.uk
CONTACT:Seyyed Ahmad Javadi
DESCRIPTION:I will present my last PhD work accepted to ACM SOCC 2019 conf
 erence. Please find the abstract below.\n\nAbstract:\nResource under-utili
 zation is common in cloud data centers. Prior works have proposed improvin
 g utilization by running provider workloads in the background\, colocated 
 with tenant workloads. However\, an important challenge that has still not
  been addressed is considering the tenant workloads as a black-box. We pre
 sent Scavenger\, a batch workload manager that opportunistically runs cont
 ainerized batch jobs next to black-box tenant VMs to improve utilization. 
 Scavenger is designed to work without requiring any offline profiling or p
 rior information about the tenant workload. To meet the tenant VMs’ reso
 urce demand at all times\, Scavenger dynamically regulates the resource us
 age of batch jobs\, including processor usage\, memory capacity\, and netw
 ork bandwidth. We experimentally evaluate Scavenger on two different testb
 eds using latency-sensitive tenant workloads colocated with Spark jobs in 
 the background and show that Scavenger significantly increases resource us
 age without compromising the resource demands of tenant VMs.
LOCATION:Computer Laboratory\, William Gates Building\, Room FW11
END:VEVENT
END:VCALENDAR