1 / 22

Cyberinfrastructure for Geospatial Computing

Cyberinfrastructure for Geospatial Computing. Bin Zhou, Chaowei Yang, Fuming Lin Joint Centre for Intelligent Spatial Computing George Mason University. Overview. Background Cyberinfrastructure Architecture GeoGrid computing platform Grid Middleware Geospatial Applications Conclusions

admon
Download Presentation

Cyberinfrastructure for Geospatial Computing

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Cyberinfrastructure for Geospatial Computing Bin Zhou, Chaowei Yang, Fuming Lin Joint Centre for Intelligent Spatial Computing George Mason University

  2. Overview • Background • Cyberinfrastructure Architecture • GeoGrid computing platform • Grid Middleware • Geospatial Applications • Conclusions • Future Work

  3. Background • Why do Geospatial Applications Need Cyberinfrastructure • Geospatial Data is Very Large • Vector data • Thousands to millions of features • Multiple dimensions • Raster data • millions of pixels (very high resolution) • Hyperspectral imagery (200+ band imagery) • Many Geospatial Applications are computing intensive

  4. Near Real-Time Requirements • Emergency response system[1] • Natural Disaster response system • Decision supporting system • Highway transportation planning system[2] • Near real-time routing system[3]

  5. Computing Characteristics • Fine-Grained • Very Short Executing Time • Huge Amount • Job Similarity • Near Real Time • Sensitive to scheduling latency • Expected Response time from seconds to several minutes

  6. They All Need • Intensive Computing Power • Huge Amount of data storage • Rapid response time • Complex & Advanced algorithms • Fast Internet Access • Interoperability & Usability • Computing problem becomes the bottle neck and poses new challenges

  7. CISC Cyberinfrastructure Architecture • CISC collaboration with SURAgrid • Siganificant Computing Power and Storage • CISC connection to LambdaRail • Fast Internet Access and collaboration • CISC Computing Pool • Fast response, powerful, interoperable and highly available

  8. CISC with SURAgrid (http://www.sura.org/programs/sura_grid.html)

  9. National Lambda Rail Full Speed Tested with WASH, ATLA, CHIC, DENV, LOSA, HOUS GMU

  10. CISC Computing Pool 224 Cores 448 G RAM

  11. CISC HPC Portal

  12. GeoGrid Computing Platform[5]

  13. Grid Middleware • Supporting Geo-spatial Computing • Condor • PBS • Lava • Globus • Other integrated middleware stack • CISC lightweight scheduling middleware (dragon) • Fast response time • Linear scheduling overhead • Efficient

  14. System Architecture Worker Central Manager User Interface Abstract Interface /APIs Services Container Algorithm module Collector Submitter Dispatcher Resource Manager Lib File Transfer Message passing Process Memory Other TCP/UDP Socket System Function

  15. Components

  16. Performances Figure 1, total finishing time with task amount[4] Figure 2,average response time with task amount [4] Figure 3 total finishing time with CPU number [4] Figure 3,average response time with task amount (real life application: near real time routing) [4]

  17. Geo-spatial Applications • Near Real Time Routing • downtown DC metro area • Dijkstra’s shortest path algorithm • time complexity of • simultaneously for 100 users • Response time 10 seconds • 478G instruction operations

  18. Regular decomposition Jibo Xie, CISC presentation, “How to grid enable geoscience applications”

  19. Test Results Effect of Different CPU Numbers and Task Amounts Average Response Time to Task Amount for 16 CPUs

  20. Conclusions • Many Geospatial applications need cyberinfrastructure support • Many grid middleware could be utilized and for some specific applications, such as fine-grained near real-time jobs, more efforts needed

  21. References [1] A Zerger and DI Smith, “Impediments to using GIS for real-time disaster decision support,” Computers, Environment and Urban Systems, pp.123-141, March 2003. [2] M Choy, MP Kwan, HV Leong, “On Real-time Distributed Geographical Database Systems,” System Sciences, pp.337-346, 1994 [3] Y. Cao “Near Real-Time Transportation Routing Supported by Grid Computing,” Ph.D. dissertation, George Mason University, Fairfax, VA, USA, pp.104 , 2007 [4] B. Zhou and C. Yang, An Effective Middleware for Fine-Grained Near Real-Time Geospatial Applications, GeoInformatica (in review), 2008. [5] Yang C., Kafatos M., Wong D., Yang R., Cao Y., 2004, GridGIS: A next generation GIS, CITSA 2004 , Jul. 21-25, Orlando, FL, pp.22-27.

  22. Thanks • Questions?

More Related