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:20200129T163603Z
LOCATION:501
DTSTART;TZID=America/Denver:20191117T091000
DTEND;TZID=America/Denver:20191117T093500
UID:submissions.supercomputing.org_SC19_sess103_ws_indis101@linklings.com
SUMMARY:SCinet DTN-as-a-Service Framework
DESCRIPTION:Workshop\n\nSCinet DTN-as-a-Service Framework\n\nYu, Chen, Mam
 bretti, Yeh, Wang...\n\nTransferring big data over Wide Area Networks (WAN
 s) is challenging because optimization is dependent on the specifics of mu
 ltiple parameters. Network services, paths, and technologies have differen
 t characteristics, including loss rate, latency, and available capacity. Y
 et, frameworks currently used to configure and orchestrate transfer system
 s, measure performance, and analyze results have limited capabilities. We 
 propose a framework, DTN-as-a-Service (DaaS), for high-performance network
  data transfers using and integration of techniques, including virtualizat
 ion, network provisioning, and performance data analysis. This framework h
 as a modular design for supporting multiple transfer tools, optimizers and
  orchestrators for the data transfer environment, including Docker and Kub
 ernetes. We present a Jupyter based workflow for high-speed network data t
 ransfer in data-intensive science and evaluate the performance of the tran
 sfer with a simple programmable visualizer implemented in the framework. W
 ith the increase in the number and the capacity of WAN links at the confer
 ences (multiple 100 Gbps WAN circuits), the challenges involved in setting
  up, testing, debugging, verifying and running applications on high-perfor
 mance systems connecting to the conference SCinet WAN circuits also increa
 se. The SCinet implementation of the DaaS framework for the conference com
 munity allowed users to control hardware, software, and network infrastruc
 ture for high-speed network data transfer, primarily for large scale appli
 cations. Through the evaluation of the framework in our test setup, we dem
 onstrated that NVMe over Fabrics with TCP is twice as efficient compared t
 o using conventional TCP in high-speed NVMe-to-NVMe transfers. We also imp
 lemented a 400 Gbps LAN experiment to evaluate the DaaS framework.\n\nTag:
  Workshop Reg Pass, Big Data, Data Analytics, Datacenter, Networks, Softwa
 re-defined networking\n\nRegistration Category: Workshop Reg Pass, Big Dat
 a, Data Analytics, Datacenter, Networks, Software-defined networking
URL:https://sc19.supercomputing.org/presentation/?id=ws_indis101&sess=sess
 103
END:VEVENT
END:VCALENDAR