Spatial Analysis on the Clouds: DEM Interpolation on the Microsoft Azure Platform Abdelmounaam Rezgui Center of Intelligent Spatial Computing for Water/Energy Science George Mason University
Spatial Cloud Computing Developing and/or Deploying Spatial Applications on Cloud Platforms Geospatial Applications 90% of all business data has a geographic component (e.g., address, sales district) Geospatial Applications Data/Compute Intensive Example: Digital Elevation Model Interpolation 2
Motivation Large volumes of data E.g., satellites collect terabytes of data daily Compute intensive algorithms Spatial analysis Short Response Time Unpredictable loads 3
Benefits of “Cloudification” Simpler Architecture On-demand Scalability Reliability Availability Maintenance cost 4
Digital Elevation Model (DEM) Interpolation Shepard's method, IDW (Inverse Distance Weighting) 5
A Previous Work by CISC Grid-based DEM Interpolation 6
Architecture & Workflow (Grid) 7
Result 8
Architecture (Cloud) 9
Benefits of “Cloudification” Simpler Architecture On-demand Scalability Reliability Availability Maintenance cost 10
DEMI on a 200-core rack Area size: ~ 80 000 sq. mi. 11
Azure Small Instance CPU: 1.6 Ghz Mem: 1.75 GB 6103 seconds Intel Core i3 CPU 540 @ 3.07 Ghz 3.07 Ghz Mem: 4.00 GB RAM 64 bit OS (Windows 7) Intel Dual Core 2.13 Ghz 2415 seconds 2.13 Ghz Mem: 4.00 GB Intel Dual Core T7100 @ 1.8 Ghz 1.8 32 bit OS (Ubuntu) Ghz 2958 seconds Mem: 2.50 GB Intel Xeon CPU 2.80 Ghz 64 bit OS (Windows Server 2008) Mem: 2.00 GB 4740 seconds 32 bit OS (CentOS) 10328 seconds 12
Azure Instance sizes 13
Amazon Instances Large Instance 7.5 GB memory m1.large 4 EC2 Compute Units (2 virtual cores with 2 EC2 Compute Units each) 850 GB instance storage 64-bit platform I/O Performance: High API name: m1.large Extra Large Instance 15 GB memory m1.xlarge 8 EC2 Compute Units (4 virtual cores with 2 EC2 Compute Units each) 1,690 GB instance storage 64-bit platform I/O Performance: High API name: m1.xlarge High-Memory Extra 17.1 GB of memory Large Instance 6.5 EC2 Compute Units (2 virtual cores with 3.25 EC2 Compute Units each) m2.xlarge 420 GB of instance storage 64-bit platform I/O Performance: Moderate API name: m2.xlarge 14
Amazon Instances High-Memory Double 34.2 GB of memory Extra Large Instance 13 EC2 Compute Units (4 virtual cores with 3.25 EC2 Compute Units each) m2.2xlarge 850 GB of instance storage 64-bit platform I/O Performance: High API name: m2.2xlarge High-Memory 68.4 GB of memory Quadruple Extra Large 26 EC2 Compute Units (8 virtual cores with 3.25 EC2 Compute Units each) Instance 1690 GB of instance storage m2.4xlarge 64-bit platform I/O Performance: High API name: m2.4xlarge High-CPU Extra Large 7 GB of memory Instance 20 EC2 Compute Units (8 virtual cores with 2.5 EC2 Compute Units each) c1.xlarge 1690 GB of instance storage 64-bit platform I/O Performance: High API name: c1.xlarge 15
Lessons Geospatial cloud computing is cost effective paradigm Computing/storage elasticity will enable new compute- and data-intensive geospatial applications Same development/deployment effort 16
Questions Thank You 17 17
Recommend
More recommend