320 likes | 467 Views
MonetDB spatial . Een open-source Spatial DBMS. Media Plaza/Het gebruik van OS software in GEO-ICT . Wilko Quak. Overzicht. TUDelft / OTB / Sectie GISt / GeoInfoNed Project De spatial DBMS markt. Categorie-en van DBMS gebruikers Intermezzo: spatial DBMS. MonetDB en MonetDB (spatial)
E N D
MonetDB spatial Een open-source Spatial DBMS Media Plaza/Het gebruik van OS software in GEO-ICT Wilko Quak
Overzicht • TUDelft / OTB / Sectie GISt / GeoInfoNed Project • De spatial DBMS markt. • Categorie-en van DBMS gebruikers • Intermezzo: spatial DBMS. • MonetDB en MonetDB (spatial) • De toekomst van MonetDB (spatial) MonetDB spatial
GIS technology MonetDB spatial
GISt visie Spatial Information Infrastructure (SII), a vision… • Heading towards an information society with varying impact on sectors of our economy • Impact on the geo-information sector is huge: geo-info and processing can be transported via (wireless) networks and can be combined with positioning • Engines of the SII nodes: geo-DBMSs (Data Base Management Systems) filled with geo-data • Fair pricing of data and services, no more endless copying of data (and loss of quality/up-to-dateness) MonetDB spatial
GISt: scientific/technological goal • Central research theme geo-DBMSas ‘glue’ between: • 3D spatio-temporal modeling • Computational geometry • Distributed GI processing • Mobile GIS (LBS) • Knowledge engineering Ambition: top 1 (3) of geo-DBMS Universities in the world MonetDB spatial
Geo Database Management Center • TU Delft open R&D center ‘geo-DBMS’ • Partnerships with ICT industry: Oracle, ESRI, Bentley • ‘Database lab’ and ‘3D GIS/VR lab’ • Website http://www.gdmc.nl for • Communication/documentation • Internet GIS MonetDB spatial
Met welk onderzoek kan de Sectie GISt bijdragen? • Onderzoek naar bruikbaarheid bestaande producten (benchmarken) • Wat mist er nog • Meedenken met ontwikkelingen • Eigen DBMS ontwikkelen als research vehikel: GeoInfoNed Project MonetDB spatial
GeoInfoNed – RGI-232 MonetDB spatial
wie • CWI – Leading DBMS experts with MonetDB • TUDelft/OTB – Knowledge of spatial processes • CycloMedia – Huge dataset and interesting problems • RWS/AGI – Large and diverse datasets and interesting problems MonetDB spatial
Wat We gaan een spatial DBMS bouwen en daar onderzoek mee doen. • Is er nog wel ruimte op de markt? Een inventarisatie: • Aanbieders • Gebruikers MonetDB spatial
De spatial DBMS markt (aanbieders) • Oracle – Marktleider commerciele segment • Informix/DB2 – Prima implemenatie. Niet erg actief • PostgreSQL/PostGIS – ‘Marktleider’ open-source • SQLServer – komt binnenkort • MySQL – open-source, beperkte functionaliteit MonetDB spatial
Spatial DBMS markt (gebuikers) 4 classes: • Server Builders: publish spatial data via web server • GIS User: Load various datasets and perform complex analyses • Data Maintainer: Maintain one core dataset • Power Users: All of the above and more MonetDB spatial
Class 1: Web Server Builders • Runs under GeoServer • You do not really need a DBMS for this (You use a fraction of DBMS functionality) • Only one query counts: Find everything within BBOX. • Multi-User • Small Queries MonetDB spatial
Class 2: GIS users • Full fledges GIS is much more powerful but SQL experts might be tempted • Main interest is functionality • Spend more time on loading data • Need a good query optimizer • Data mining MonetDB spatial
Class 3: Dataset Maintainers • Limited number of queries. • 24/7 stability • Transactions. • Clustering of data after updates is interesting. MonetDB spatial
Class 4: Power users • All of the above and more • Take time to understand software • Run into research problems: • Point clouds from laser-scanning • Support 3d-volumes • Realtime dynamic data MonetDB spatial
Intermezzo: Spatial DBMS • Spatial data types • Operaties op types • Indexeren op types MonetDB spatial
Spatial Data Types • Basic types: Point Line Polygon • More Complex: 3D, Geodetic Coordinates MonetDB spatial
Spatial Data types volgens OpenGIS Point LineString LinearRing Polygon MultiPoint MultiLineString MultiPolygon MonetDB spatial
Functions op spatial types MonetDB spatial
D E H A F G I J L M K B C N A B C Branching factor 4 L M N D E F G H I J K Indexeren spatial types • R-tree puts MBR on object • MBR : minimum bounding rectangle • Object can be of type point, line or region • At lowest (leaf) level objects are grouped to larger MBR’s, which can be grouped (and so on .....) Einde Intermezzo MonetDB spatial
MonetDB Introduction* MonetDB is opens • Hardware trends • MonetDB design considerations • MonetDB architecture *Slides borrowed from CWI MonetDB spatial
Hardware Trends 50% p/year: - cpu speed - mem size - mem bandwidth - disk bandwidth 1% p/year: - mem latency 10% p/year: - disk latency
Latency is the enemy! • Commercial DBMS products (oracle, DB2, SQLserver) stem from OLTP roots • focus on minimizing random I/Os => depend on latency! • MonetDB: built for bulk access • optimize CPU and memory performance
MonetDB design considerations • Multi-model database kernel support • Extensible data types, operators, accelerators • Database hot-set is memory resident • Simple data structures are better • Index management should be automatic • Do not replicate the operating system • Optimize when you know the situation • Cooperative transaction management MonetDB spatial
MonetDB product family End-user application XQuery SQL PHP JDBC ODBC Python Perl C-mapi lib MAPI protocol Monet kernels MonetDB spatial
strength weakness Triggers and SQL procs missing; XQuery; some overhead in simple SQL queries; MIL scripting; missing GIS module State-of-the-art SQL & XQuery DBMS; multi-model database support; decomposed storage model; query processing for decision support systems;extensible kernel; multiple APIs;….. opportunities threats Commercial systems requirements; competition established market; floating interest; development skills and quality awareness Unique selling points for advanced applications in multimedia, streaming, covering both small and the large world SWOT GET INSPIRED BY ADVANCED APPLICATIONS MonetDB spatial
MonetDB - Physical data organization • Binary Association Tables MonetDB spatial
Spatial MonetDB MonetDB spatial
Spatial MonetDB -- nu • Eerste implementatie in volgende release (Q4). • OpenGIS implementatie gebaseerd op PostGIS • OpenJUMP plugin • Shape2MonetDB converter • GeoTools Plugin (in ontwikkeling) MonetDB spatial
Spatial MonetDB – de toekomst • MonetDB moet het hebben van zijn speciale architectuur. • Uitdagingen: • Puntenwolken van Laserscanner. • Data-minen op grote spatio-temporele datasets. • XML-Data. • Saaier werk: • Testen huidige implementatie • debuggen MonetDB spatial
VragenDiscussie MonetDB spatial