1 / 45

James Mooney, Brett Morgan, Kevin Ross, Adam Ruf, Ben Sigrist, and Art Lembo

Freely available enterprise-class spatial database servers: experiences with Oracle 10g Express, SQLServer Express 2008 (CTP), and PostGIS. James Mooney, Brett Morgan, Kevin Ross, Adam Ruf, Ben Sigrist, and Art Lembo. Introduction.

walden
Download Presentation

James Mooney, Brett Morgan, Kevin Ross, Adam Ruf, Ben Sigrist, and Art Lembo

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. Freely available enterprise-class spatial database servers: experiences with Oracle 10g Express, SQLServer Express 2008 (CTP), and PostGIS James Mooney, Brett Morgan, Kevin Ross, Adam Ruf, Ben Sigrist, and Art Lembo

  2. Introduction • Purpose: chronicle experiences in using free offerings of spatial database management systems for GIS. • Installation, loading, analysis, and display • Databases • Oracle Spatial Express • Microsoft SQLServer Express 2008 (CTP) • PostGRES/PostGIS

  3. Background • The traditional GIS marketplace is rapidly changing. • Google: One of the largest Internet-based content providers introduces Google Earth and Google Maps, introducing geospatial information to the mass market. • Microsoft: World’s largest software company introduces spatial database capabilities in their product offerings, and introduces Microsoft Virtual Earth to compete with Google maps. Spatial capability is now available to all Microsoft Office users. • Open Source: A burgeoning market of freely developed geospatial software places sophisticated geospatial technology in the hands non-traditional users.

  4. Background • Results of a new geospatial information super-highway • Volume: more users are now aware of a geospatial information super-highway. • Affordability: more users are now able to afford the tools to drive on the geospatial information super-highway • Navigation: more users require assistance to utilize the tools of the geospatial information super-highway • What geospatial information super-highway capabilities are available to the general public at little or no cost, and what is the potential for using these tools?

  5. Spatial databases PostGIS SQLServer Express Oracle Express

  6. Oracle • Oracle Database 10g Express Edition • Current Version 10.2.0.1.0 (Express Edition) • 32-bit • Company’s Purpose: • http://www.oracle.com/technology/products/database/xe/index.html Companies • Microsoft • Microsoft: Redman, WA • Microsoft is a software company that has widespread adoption in the marketplace • 6 Editions of software from Compact Edition to Enterprise Edition. • Prices range from $0 to $25,000/processor • We tested the Express Edition • Latest version: SQL Server 2008 • PostGIS • PostGIS is a spatial extension to PostgreSQL, an advanced open source database • Developed and maintained by Refractions Research • Freely available, Enterprise grade, Active user community • http://postgis.refractions.net/

  7. Supported Data Types SQLSERVER • Proprietary data format • Conforms to OGC standards • WKT (Well Known Text) • Geometry Data Type • Represents data in an Euclidean (flat) coordinate system • Geography Data Type • Represents data in an Ellipsoidal (round-earth) coordinate system • Examples of object types • Point • MultiPoint • LineString • MultiLineString • Polygon • MultiPolygon • GeometryCollection POSTGIS • OpenGIS Consortium Standards • WKT (Well Known Text) • WKB (Well Known Binary) • Both of these formats include information about the type of the object as well as the coordinates which form the object • Examples of object types • POINT (0 0) • LINESTRING (0 0,1 3,5 7) • POLYGON ((0 0,1 1,2 2,2 0,0 0),(0 5,5 1,6 1,0 5) • MULTIPOINT (0 0,1 3) • MULTILINESTRING ((0 0,1 1,1 2),(2 3,3 2,5 4)) • MULTIPOLYGON(((0 0,4 0,4 4,0 4,0 0),(1 1,2 1,2 2,1 2,1 1)), ((-1 -1,-1 -2,-2 -2,-2 -1,-1 -1))) • GEOMETRYCOLLECTION(POINT(2 3),LINESTRING((2 3,3 4))) ORACLE • OpenGIS Consortium Standards • WKT (Well Known Text) • WKB (Well Known binary) • Oracle spatial

  8. Known Limitations SQLServer Data type restrictions Each Geog instance must fit in a Hemisphere Instance from (OGC), (WKT),(WKB) object larger than a hemisphere throws an Argument exception Geog methods that require input from two Geog instances (intersect,union,differences,symdiff) return null if bigger than a single hemisphere Oracle • Memory – Only 1gb RAM is addressed, limiting the number of concurrent users. • XE will only use one CPU. It can be run on a multi-CPU configuration, but will not scale up to use those other CPUs. • Only one XE Database may be run on an individual computer. • 4gb Disk Space limitation.

  9. Architecture for testing • Hardware • Standard PC with Microsoft XP • Shuttle PC P2 Prima 3500 • 64-bit, 3.0 GHZ. • 8 GB RAM • Quad-core • Software • Spatial databases: Oracle, SQLServer, • PostGIS • Viewing: Manifold GIS, Quantum GIS • Loading: Manifold GIS, SPIT, SharpNet • Data • Simple point, line, area • City of Ithaca parcels (5,000 area features) • Large fictitious grid (5 million area features) • TIGER data for Northeast US (4 million linear features)

  10. Installation

  11. Installation • Each product has a Window’s based installation process • Oracle had the simplest installation process • PostGIS and SQLServer had numerous difficulties within the University network. • PostGIS and SQLServer worked fine on individual computers

  12. Loading Spatial Data

  13. Simple line insertion with SQL

  14. Line Segment

  15. Oracle Data Loading

  16. Data File

  17. Ithparc2 Linked to Oracle

  18. Query in SQLServer sqlcmd

  19. The query in manifold along with the original parcel drawing and the one from the database.

  20. Loading data with 3rd party products

  21. Manifold’s database console was used to Connect to the new database.

  22. SQLServer - Exporting data with Manifold - Exporting data using Manifold is fairly straightforward. - Used Manifold to export BigBadData (5 million records) and the roads data (4 million records) into separate databases • The BigBadData took about one hour and twenty minutes • The roads data took about forty five minutes

  23. SharpGIS Shapefile uploader • This tool was recently created by Morten Nielsen • It can be obtained freely from http://www.sharpgis.net • This tool was designed to easily upload ESRI Shapefiles into SQLServer 2008

  24. This shows the County Data Opened in manifold

  25. PostGIS with SPIT

  26. Traditional and Spatial Queries

  27. SQL Queries SQLServer • SELECT * From dbo.Cnty_sf1_Drawing where P001001 > 100000 • SELECT SUM(P001001) as SUM FROM dbo.Cnty_sf1_Drawing • SELECT AVG(P001001) as AVG FROM dbo.Cnty_sf1_Drawing PostGIS • SELECT * FROM "BigBadData” WHERE "Population" > 90000 • SELECT SUM("Population") AS Sum FROM "BigBadData" • SELECT AVG("Population") AS Average FROM "BigBadData" Oracle • SELECT GEODESC FROM Cnty_sf1_drawing WHERE P001001 > 100000; • SELECT SUM(P001001) AS SUM FROM Cnty_sf1_drawing; • SELECT AVG(P001001) AS AVG FROM Cnty_sf1_drawing;

  28. SQLServerSpatial Queries Buffer DECLARE @Wicomico geometry; SET @Wicomico= (SELECT top 1 geom FROM dbo.Cnty_sf1 WHERE GEODESC= 'Wicomico County'); SET @Wicomico= @Wicomico.MakeValid(); SET @Wicomico= (SELECT @Wicomico.STBuffer(250).ToString()); Insert into bobo2(geom) VALUES(@Wicomico);

  29. SQLServerSpatial Queries UNION DECLARE @Parc1 geometry; DECLARE @Parc2 geometry; DECLARE @Result geometry; SET @Parc1= (SELECT top 1 geom1 FROM dbo.Parcels WHERE gid= 27); SET @Parc2= (SELECT top 1 geom1 FROM dbo.Parcels WHERE gid= 419); SET @Parc1= @Parc1.MakeValid(); SET @Parc2= @Parc2.MakeValid(); SET @Result= (SELECT @Parc1.STUnion(@Parc2)); INSERT INTO Dist(geom) VALUES (@Result);

  30. Oracle Spatial Query: Buffer SDO_BUFFER(Geometry, Distance in meters, Tolerance)

  31. SDO_UNION(Geometry1, Geometry2, Tolerance) Oracle Spatial Query: Union

  32. PostGISSpatial Queries BUFFER SELECT ST_Buffer(geometry,.1) FROM "BigBadData" WHERE "Population" > 90000

  33. PostGISSpatial Queries UNION SELECT ST_UNION((SELECT geometry FROM "Parcels" WHERE gid = 27),(SELECT geometry FROM "Parcels" WHERE gid = 419))

  34. Discussion Observations and expectations of first time users

  35. PostGIS • Positive • SQL syntax was the most straightforward • PGAdmin was very easy to use (little directions necessary) • Useful GUI capabilities for management • Easy integration with Manifold and QGIS • Large and helpful user committee • Negative • Installation in certain environments was difficult • Very large datasets locked up the server (possibly due to our hardware configuration)

  36. Oracle • Positive • Very easy installation • Good documentation • Spatial SQL easy to write • Administration tool easy to use • Easy integration with Manifold • Negative • Oracle Express limited character length in enormous INSERT statements • Geometry queries required more thought, as more parameters were required. • Use of control files was cumbersome

  37. SQLServer Express 2008 • Must remember that this was a CTP, and part of the objective is to uncover inefficiencies. Therefore, this is not a criticism of SQLServer Express 2008. • Positive • Good online help • SQL straightforward • Relatively fast processing • Good integration with Manifold and SharpGIS • Negative • Installation was difficult (but remember, it was a CTP) • Spatial SQL more cumbersome than other products • Currently no management studio • Reverse format of X,Y coordinates made certain loading difficult.

  38. Future Activities • GEOG 415 – Special topics class (Spring 2009) at Salisbury University in building enterprise spatial database management systems • Currently evaluating pilot projects for implementing an enterprise-wide, low cost spatial database management system within a local Eastern Shore agency. • EFRI- RESIN: Decision Support for Water Supplies and Critical Interdependent Infastructure • Central enterprise coordination of shared geospatial storage (Cornell University, University of Delaware, University of Southern California, and Salisbury University) for a large National Science Foundation (NSF) proposal. Spatial database includes PostGRES/PostGIS with client access from ArcGIS, Manifold GIS, Quantum GIS, and Microsoft Office applications.

  39. Conclusion • Many free versions of large-scale spatial databases exist beyond what we investigated • Products were easy enough to use so that an evaluation of all three was possible in a one semester course • Each software product had its strengths and weaknesses in terms of • Installation • Data loading • Data analysis • Data integration with third party products • Having a third party product to manage data loading, analysis and display is beneficial • Cutting one’s teeth on the free versions of the data products are an ideal way to prepare a full migration to the full package

  40. Workgroup ideas • Manifold does not really operate in a true client/server implementation with spatial databases • Still too committed to caching all the data • What are the tradeoffs of this approach? • Should Manifold become a true client to these spatial databases? • Should Manifold create their own middle-ware tier? • What are the tradeoffs of this approach?

More Related