ames research center nas division 1 experience with nasa’s grid miner thomas h. hinke nasa ames...
TRANSCRIPT
Ames Research Center NAS Division1
Experience With NASA’s Grid Miner
Thomas H. Hinke
NASA Ames Research Center
Moffett Field, California, USA
Ames Research Center NAS Division2
Outline
• Why use the grid for data mining?
• Overview of Grid Miner
• Experience adapting existing stand-along miner to grid
• A recent application of the Grid Miner
Ames Research Center NAS Division3
Grid Provides Computational Power
• Grid couples needed computational power to data– NASA has a large volume of data stored in its
distributed archives• E.g., In the Earth Science area, the Earth Observing System
Data and Information System (EOSDIS) holds large volume of data at multiple archives
– Data archives are not designed to support user processing
– Grids, coupled to archives, could provide such a computational capability for users
Ames Research Center NAS Division4
Grid Provides Re-Usable Functions• Grid-provided functions do not have to be re-implemented
for each new mining system– Single sign-on security – Ability to execute jobs at multiple remote sites– Ability to securely move data between sites– Broker to determine best place to execute mining job– Job manager to control mining jobs
• Mining system developers do not have to re-implement common grid services
• Mining system developers can focus on the mining applications and not the issues associated with distributed processing
Ames Research Center NAS Division5
Grid Will Provide Re-usable Services
• In the future, Grid/Web services will provide the ability to create reusable services that can facilitate the development of data mining systems– Builds on the web services work from the e-
commerce area• Service interface is defined through WSDL (Web
Services Description Language)• Standard access protocol is SOAP (Simple Object
Access Protocol)
Ames Research Center NAS Division6
Grid Services: A Foundation for Grid Mining
• Global Grid Forum working groups on – Open Grid Services Architecture (OGSA) standard
under development to specify a grid-enabled web services architecture. See “Physiology of the Grid: An Open Grid Services Architecture for Distributed Systems Integration”
– Open Grid Services Infrastructure (OGSI) standard has been released. Specifies common interfaces that all grid services should support.
Ames Research Center NAS Division7
Grid Mining and OGSA/OGSI
• An OGSA/OGSI compliant mining service could be build
• Mining applications could be built by re-using capabilities provided by existing grid services.
Ames Research Center NAS Division8
Outline
• Why use the grid for data mining?
• Overview of Grid Miner
• Experience adapting existing stand-along miner to grid
• A recent application of the Grid Miner
Ames Research Center NAS Division9
Grid Miner
• Developed as one of the early applications on the IPG– Helped debug the IPG
– Provided basis for satisfying a major IPG milestones
• IPG is NASA implementation of Globus-based Grid
• Provides basis for what could be an on-going Grid Mining Service
Grid Miner OperationsGrid Miner Operations Preprocessed
DataPreprocessed
DataDataData
TranslatedData
Patterns/ModelsPatterns/Models
ResultsResults
OutputGIF ImagesHDF-EOSHDF Raster ImagesHDF SDSPolygons (ASCII, DXF)SSM/I MSFC
Brightness TempTIFF ImagesOthers...
Preprocessing AnalysisClustering K Means Isodata MaximumPattern Recognition Bayes Classifier Min. Dist. ClassifierImage Analysis Boundary Detection Cooccurrence Matrix Dilation and Erosion Histogram Operations Polygon Circumscript Spatial Filtering Texture OperationsGenetic AlgorithmsNeural NetworksOthers...
Selection and Sampling Subsetting Subsampling Select by Value Coincidence SearchGrid Manipulation Grid Creation Bin Aggregate Bin Select Grid Aggregate Grid Select Find HolesImage Processing Cropping Inversion ThresholdingOthers...
InputHDFHDF-EOSGIF PIP-2SSM/I PathfinderSSM/I TDRSSM/I NESDIS Lvl 1BSSM/I MSFC
Brightness TempUS RainLandsatASCII GrassVectors (ASCII Text)
Intergraph RasterOthers...
Figure thanks to Information and Technology Laboratory at the University of Alabama in Huntsville
Ames Research Center NAS Division11
Mining on the Grid
Grid Mining Agent
IPG Processor
Satellite Data
Archive X
Satellite Data
Archive Y
Grid Mining Agent
IPG Processor
Grid Mining Agent
IPG Processor
Ames Research Center NAS Division12
Grid Miner Architecture
Grid Mining Agent
IPG Processor
MiningDatabaseDaemon
ControlDatabase
IPG Processor
Grid Mining Agent
IPG Processor
Mining OperationsRepository
IPG Processor
Data
Archive X
Satellite Data
Archive Y
MinerConfiigServer
IPG Processor
Ames Research Center NAS Division13
Example: Mining for Mesoscale Convective Systems
Image shows results from mining SSM/I data
Ames Research Center NAS Division14
Outline
• Why use the grid for data mining?
• Overview of Grid Miner
• Experience adapting existing stand-along miner to grid
• A recent application of the Grid Miner
Ames Research Center NAS Division15
Starting Point for Grid Miner• Grid Miner reused code from object-oriented ADaM data
mining system – Developed under NASA grant at the University of Alabama in
Huntsville, USA– Implemented in C++ as stand-alone, objected-oriented mining
system• Runs on NT, IRIX, Linux
– Has been used to support research personnel at the Global Hydrology and Climate Center and a few other sites.
• Object-oriented nature of ADaM provided excellent base for enhancements to transform ADaM into Grid Miner
Ames Research Center NAS Division16
Transforming Stand-Alone Data Miner into Grid Miner
• Original stand-alone miner had 459 C++ classes.
• Had to make small modifications to ADaM– Modified 5 existing classes – Added 3 new classes
• Grid commands added for– Staging miner agent to remote sites– Moving data to mining processor
Ames Research Center NAS Division17
Staging Data Mining Agent to Remote Processor
globusrun -w -r target_processor '&(executable=$(GLOBUSRUN_GASS_URL)# path_to_agent)(arguments=arg1 arg2 … argN)(minMemory=500)'
Ames Research Center NAS Division18
Moving Data to be Mined
gsincftpget remote_processor local_directory remote_file
Ames Research Center NAS Division19
Outline
• Why use the grid for data mining?
• Overview of Grid Miner
• Experience adapting existing stand-along miner to grid
• A recent application of the Grid Miner
Ames Research Center NAS Division20
Demonstrate Grid Support for Interdisciplinary Earth Science Research
• Goal: Combine data from two distinctly different instruments (stored on two different grid-connected mass storage systems) to produce new insights by looking at data covering the same time and place across data from the two different instruments.
• Approach: – Use Grid Miner to mine TMI data for mesoscale convective
systems.– Generate feature index (convex hull polygon) for all
mesoscale convective systems found.– Transmit polygons in form of XML document to subsetter. – Subset CERES SSF data that corresponds to mesoscale
convective systems discovered by Grid Miner
Ames Research Center NAS Division21
Desired Processing Pattern
Data Archive
Data Archive Network Subsetting
Data Mining
NASA Atmospheric Sciences Data Center
Data Archived at NASA Ames
Grid Processing
To User
Ideally Grid Miner would use grid resources co-located with the data, but if not, could use
available remote grid resources
The Details
Data Cache
Grid Processor
CERESData
Archive
TMI Data on Mass Store
Storage Resource Broker
Grid Mining Agent
IPG Processor
Subsetter
Grid Processor
LaRC Atmospheric Sciences Data Center
IPG
MSCP (2) sends mining plan and (5) retrievesFeature Index
(MSCP) Miner-SubsetterControl Program
Grid Processor
Miner Executables
IPG Processor
IPG Broker
IPG Processor
(1) Broker Selects IPG Resource for Mining
(3) MSCP transfers GridAgent to Mining Site using Job Manager (not shown)
(4) GridAgent Transfers Mining Ops
(6) MSCP Starts Subsetter on Feature Index
(5)
Ames Research Center NAS Division23
Example of Data Being Mined
• 230 MB contained in 15 orbit files for one day of TMI (TRMM [Tropical Rainfall Measuring Mission] Microwave Imager) data
• Much higher resolution data exists with significantly higher volume.
Mining and Subsetting Results
QuickTime™ and aTIFF (Uncompressed) decompressorare needed to see this picture.
Grid Miner produced XML (Extensible Markup Language) document of polygons that circumscribe mesoscale convective systems. (MCSs) The following shows a portion of the XML description for two of the 64 vertices that comprise the convex hull polygon produced for the third MCS found by the miner in TMI data for April 1, 1998:
<polygon><julian_date_time> 2450904.754815 </julian_date_time><human_date_time> 1998-04-01 GMT 06:06:56 </human_date_time><size_in_square_km> 2083.126221 </size_in_square_km><region_type> 2 </region_type><vertices><number_of_vertices> 64 </number_of_vertices><vertex><latitude> -2.26 </latitude><longitude> -178.28 </longitude></vertex><vertex><latitude> -2.08 </latitude><longitude> -178.38 </longitude></vertex>...</polygon></polygon_list>
Grid Miner produced view of area mined using TMI data.
CERES SSF footprints for April 1, 1998 hour 6 corresponding to the third MCS found by Grid MinerConvex Hull: 3 with 9
Footprint: 1 15804Footprint: 2 15805Footprint: 3 16090Footprint: 4 16091Footprint: 5 16094Footprint: 6 16376Footprint: 7 16377Footprint: 8 16381Footprint: 9 16382
Ames Research Center NAS Division25
Current Status• Currently works on the IPG as a prototype system• User documentation underway• Data archives need to be grid-enabled
– Connected to the grid– Provide controlled access to data on tertiary storage
• E.g., by using a system such as the Storage Resource Broker that was developed at the San Diego Super Computer Center
• Some earlier-adopter users need to be found to begin using the Grid Miner– Willing to code any new operations needed for their
applications– Willing to work with system with prototype-level
documentation