CloudCast: Cloud Computing for Short-term Mobile Weather Forecasts Dilip Kumar Krishnappa, David Irwin, Eric Lyons and Michael Zink IPCCC 2012.

Slides:



Advertisements
Similar presentations
Sponsored by the National Science Foundation GENI I&M Workshop NetCDF and Local Data Manager (LDM) Mike Zink November 4, 2010
Advertisements

SLA-Oriented Resource Provisioning for Cloud Computing
Data-Intensive Cloud Control for GENI GEC 8 demo Orca control framework July 20, 2010 Michael Zink, Prashant Shenoy, Jim Kurose, David Irwin and Emmanuel.
Sponsored by the National Science Foundation GENI Alpha Demonstration Nowcasting: UMass/CASA Weather Radar Demonstration David Irwin November 3, 2010
Deploying GMP Applications Scott Fry, Director of Professional Services.
Chapter 4 Infrastructure as a Service (IaaS)
Virtual Machine Usage in Cloud Computing for Amazon EE126: Computer Engineering Connor Cunningham Tufts University 12/1/14 “Virtual Machine Usage in Cloud.
Webscale Computing Mike Culver Amazon Web Services.
1. Topics Is Cloud Computing the way to go? ARC ABM Review Configuration Basics Setting up the ARC Cloud-Based ABM Hardware Configuration Software Configuration.
Proactive Prediction Models for Web Application Resource Provisioning in the Cloud _______________________________ Samuel A. Ajila & Bankole A. Akindele.
Click to add text Introduction to the new mainframe: Large-Scale Commercial Computing © Copyright IBM Corp., All rights reserved. Chapter 3: Scalability.
What should you Cache? A Global Analysis on YouTube Related Video Caching Dilip Kumar Krishnappa, Michael Zink and Carsten Griwodz NOSSDAV 2013.
Aneka: A Software Platform for .NET-based Cloud Computing
Energy Efficient Prefetching – from models to Implementation 6/19/ Adam Manzanares and Xiao Qin Department of Computer Science and Software Engineering.
Energy Efficient Prefetching with Buffer Disks for Cluster File Systems 6/19/ Adam Manzanares and Xiao Qin Department of Computer Science and Software.
1 Exploring Data Reliability Tradeoffs in Replicated Storage Systems NetSysLab The University of British Columbia Abdullah Gharaibeh Matei Ripeanu.
INTRODUCTION TO CLOUD COMPUTING CS 595 LECTURE 4.
A Day in the Life of an Application Performance Engineer Keith Lyon - Shunra Software
Bandwidth Measurements for VMs in Cloud Amit Gupta and Rohit Ranchal Ref. Cloud Monitoring Framework by H. Khandelwal, R. Kompella and R. Ramasubramanian.
Sponsored by the National Science Foundation GENI Alpha Demonstration Nowcasting: UMass/CASA Weather Radar Demonstration Mike Zink, David Irwin LEARN Workshop,
Cloud computing Tahani aljehani.
Exploiting Virtualization for Delivering Cloud based IPTV Services Speaker : 吳靖緯 MA0G IEEE Conference on Computer Communications Workshops.
Running Your Database in the Cloud Eran Levin VP R&D - Xeround.
CERN IT Department CH-1211 Genève 23 Switzerland t Next generation of virtual infrastructure with Hyper-V Michal Kwiatek, Juraj Sucik, Rafal.
Introduction to Cloud Computing
A User Experience-based Cloud Service Redeployment Mechanism KANG Yu.
A measurement study of vehicular internet access using in situ Wi-Fi networks Vladimir Bychkovsky, Bret Hull, Allen Miu, Hari Balakrishnan, and Samuel.
PhD course - Milan, March /09/ Some additional words about cloud computing Lionel Brunie National Institute of Applied Science (INSA) LIRIS.
Multimedia and Mobile communications Laboratory Augmenting Mobile 3G Using WiFi Aruna Balasubramanian, Ratul Mahajan, Arun Venkataramani Jimin.
CLOUD COMPUTING 2.0 Finally, the promise of the cloud has arrived v 1.8.
Additional Areas Mike Zink CASA Deputy Director University of Massachusetts NSF Year 9 Visit, July 2 nd, 2012.
Sponsored by the National Science Foundation Nowcasting: UMass/CASA Weather Radar Demonstration Michael Zink CC-NIE Workshop January 7, 2013.
Cloud Computing. What is Cloud Computing? Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable.
Sponsored by the National Science Foundation GENI and Cloud Computing Niky RIga GENI Project Office
Ocean Observatories Initiative Common Execution Infrastructure (CEI) Overview Michael Meisinger September 29, 2009.
Presented by: Sanketh Beerabbi University of Central Florida COP Cloud Computing.
An Autonomic Framework in Cloud Environment Jiedan Zhu Advisor: Prof. Gagan Agrawal.
Cansys West International Conference February , 2013Panama City, Panama An easier way to deliver APPX applications.
Department of Information Engineering The Chinese University of Hong Kong A Framework for Monitoring and Measuring a Large-Scale Distributed System in.
BFTCloud: A Byzantine Fault Tolerance Framework for Voluntary-Resource Cloud Computing Yilei Zhang, Zibin Zheng, and Michael R. Lyu
Grids, Clouds and the Community. Cloud Technology and the NGS Steve Thorn Edinburgh University Matteo Turilli, Oxford University Presented by David Fergusson.
1 A Framework for Data-Intensive Computing with Cloud Bursting Tekin Bicer David ChiuGagan Agrawal Department of Compute Science and Engineering The Ohio.
The NWS/NCAR “Forecaster Over the Loop” Fort Worth Operational Demonstration Human Enhancement of a Thunderstorm Nowcasting System Eric Nelson, Rita Roberts,
An Architecture for Distributed High Performance Video Processing in the Cloud 作者 :Pereira, R.; Azambuja, M.; Breitman, K.; Endler, M. 出處 :2010 IEEE 3rd.
Looking Ahead: A New PSU Research Cloud Architecture Chuck Gilbert - Systems Architect and Systems Team Lead Research CI Coordinating Committee Meeting.
The New Zealand Institute for Plant & Food Research Limited Use of Cloud computing in impact assessment of climate change Kwang Soo Kim and Doug MacKenzie.
Sponsored by the National Science Foundation University of Massachusetts Amherst November 2 nd, 2011 GENI DiCloud.
V I SE/D I C LOUD S TATUS J ULY 28 TH, 2011 Michael Zink ECE Department University of Massachusetts Amherst.
CCGrid, 2012 Supporting User Defined Subsetting and Aggregation over Parallel NetCDF Datasets Yu Su and Gagan Agrawal Department of Computer Science and.
Near Real-Time Verification At The Forecast Systems Laboratory: An Operational Perspective Michael P. Kay (CIRES/FSL/NOAA) Jennifer L. Mahoney (FSL/NOAA)
A User Experience-based Cloud Service Redeployment Mechanism KANG Yu Yu Kang, Yangfan Zhou, Zibin Zheng, and Michael R. Lyu {ykang,yfzhou,
Data-Intensive Cloud Control for GENI GEC 10 Orca control framework March 15 th, 2011 Michael Zink, Prashant Shenoy, Jim Kurose, David Irwin and Emmanuel.
Computing Research Testbeds as a Service: Supporting large scale Experiments and Testing SC12 Birds of a Feather November.
Aneka Cloud ApplicationPlatform. Introduction Aneka consists of a scalable cloud middleware that can be deployed on top of heterogeneous computing resources.
3/12/2013Computer Engg, IIT(BHU)1 CLOUD COMPUTING-1.
1 Querying the Physical World Son, In Keun Lim, Yong Hun.
Authors: Jiann-Liang Chenz, Szu-Lin Wuy, Yang-Fang Li, Pei-Jia Yang,
Cloud Computing Shannon McManus Michael Weihert. What is Cloud Computing?
Euro-Par, HASTE: An Adaptive Middleware for Supporting Time-Critical Event Handling in Distributed Environments ICAC 2008 Conference June 2 nd,
Smart Grid Big Data: Automating Analysis of Distribution Systems Steve Pascoe Manager Business Development E&O - NISC.
KAASHIV INFOTECH – A SOFTWARE CUM RESEARCH COMPANY IN ELECTRONICS, ELECTRICAL, CIVIL AND MECHANICAL AREAS
© 2012 Eucalyptus Systems, Inc. Cloud Computing Introduction Eucalyptus Education Services 2.
Accelerate your ambition Partner Billing and Reporting.
Hao Wu, Shangping Ren, Gabriele Garzoglio, Steven Timm, Gerard Bernabeu, Hyun Woo Kim, Keith Chadwick, Seo-Young Noh A Reference Model for Virtual Machine.
Nowcasting: UMass/CASA Weather Radar Demonstration David Irwin
WP18, High-speed data recording Krzysztof Wrona, European XFEL
Diskpool and cloud storage benchmarks used in IT-DSS
Bandwidth Measurements for VMs in Cloud
Vlad Nae, Radu Prodan, Thomas Fahringer Institute of Computer Science
User interaction and workflow management in Grid enabled e-VLBI experiments Dominik Stokłosa Poznań Supercomputing and Networking Center, Supercomputing.
Presentation transcript:

CloudCast: Cloud Computing for Short-term Mobile Weather Forecasts Dilip Kumar Krishnappa, David Irwin, Eric Lyons and Michael Zink IPCCC 2012

2 ECE Department Outline  Introduction  Cloud Services  CloudCast Architecture  Cloud Measurements  Nowcasting  Conclusion

3 ECE Department Motivation In 75 days, 90 hours of convective precipitation Nowcast only useful during these events Cost for dedicated hardware $4000 (does not include power, AC, maintenance) Idle for 95% of the time At 45 cents/ hour, 90 hours of active use would cost $40 per product, plus $2 per user to stream the resultant data out of the cloud

4 ECE Department Nowcast Example

5 ECE Department Introduction  Dedicating high-end servers for intermittent requirement wastes resources.  Cloud services IaaS model suits such applications.  Short term weather forecasting is one such real-time application.  Cost of operation for weather predicted last year for 75 days shows 50$ of cloud usage vs. 4000$ of dedicated hardware.  Are cloud services well suited for such applications?

6 ECE Department Introduction (Contd..)  We consider real-time application of short-term weather forecasting called “Nowcasting”.  Nowcasting produces short- term forecasts for 10s of minutes in the future.  Nowcasting has strict time constraints.  Are cloud services capable of keeping up with such strict time constraints?

7 ECE Department Cloud Services  Commercial Cloud Services: Amazon EC2, Rackspace  Research Cloud Testbeds: GENICloud, ExoGENI  Commercial cloud services provide on-demand resources based on per hour pricing.  Pricing depends on the instance requested and the amount of I/O performed per hour.

8 ECE Department Cloud Services (Contd..)  GENICloud is an open source cloud platform using Eucalyptus as a base.  GENICloud provides slice-based architecture to acquire cloud instances.  ExoGENI is open source cloud platform provisioned by ORCA control framework.  ExoGENI provides flexibility to users to upload the image of their own and extend the size of the image.

9 ECE Department NEXRAD Radars

10 ECE Department Measurements  PlanetLab measurement to mimic the NEXRAD radars.  159 NEXRAD radar sites around the country. 103 PlanetLab nodes found near to those locations.  According to our own radar testbed setup (CASA), 5 Mbps is the minimum throughput of data transfer from radars.

11 ECE Department Measurements (Contd..)  Three types of data transfer experiment. - Serial Data Transfer - Parallel Data Transfer - Distributed Data Transfer  Parallel transfer helps to understand the bottleneck in the system.  Data is transferred to nearest instance based on the subset of nodes in distributed data ingest.

12 ECE Department Distributed Data transfer Radar2 Radar1Radar3 Radar4 Cloud Instance Cloud Instance1Cloud Instance2 Radar1 Radar2 Radar3 Radar4 Parallel Data Transfer Distributed Data Transfer

13 ECE Department Results – Serial, Commercial Cloud

14 ECE Department Results – Serial, Research Cloud

15 ECE Department Summary of Results Measurement Type EC2 East (Mbps) EC2 West (Mbps) Rackspace (Mbps) GENICloud (Mbps) ExoGENI (Mbps) Serial Parallel Distributed

16 ECE Department Computation Time Analysis Instance Type Memory (GB) Disk (GB) CPUCost/ hr ($) Total Cost ($) Exec. Time (sec) Total Time (sec) Amzon EC Rackspace GENICloud ExoGENI

17 ECE Department Nowcasting  Main objective of our work is the feasibility of Cloud Services for short-term weather forecasting.  We have found from our studies that, cloud services are feasible for Nowcasting.  We performed live measurement and analysis to determine the overall time taken to generate 15-minute Nowcasts in each of the cloud services considered.  We find that it takes just about 2 minutes max for the generation of 15-minute Nowcasts and sending the results to the end user.

18 ECE Department Data Aggregation

19 ECE Department DFW Radar Deployment NETWORK ROLLOUT: First phase in blue, Second phase in green UNT UTA Town of Addison City of Fort Worth

20 ECE Department Installation of UTA radar

21 ECE Department Future Scenario

22 ECE Department Conclusion  In our work, we look into the feasibility of cloud services for real time scientific applications.  Scientific application in our case is short-term weather forecasting called Nowcasting.  We perform a series of measurements to determine our objective.  From our measurements, we find that cloud services perform best when there is less number of radar nodes sending data to a single instance.  We conclude that, having a subset of radar nodes sending data to a single instance works best for our real-time application.

23 ECE Department US Ignite – Ultra-high Bandwidth Operations Control Center Short-term Forecast in the Cloud Short-term Forecast in the Cloud Long-term, Large-scale Forecast Long-term, Large-scale Forecast End Users: Public, NWS EM,Media Industry End Users: Public, NWS EM,Media Industry Up to 150 Mbps Up to n * 150 Mbps