Describing and Discovering Mathematical Web and Grid Services Mike - - PowerPoint PPT Presentation

describing and discovering mathematical web and grid
SMART_READER_LITE
LIVE PREVIEW

Describing and Discovering Mathematical Web and Grid Services Mike - - PowerPoint PPT Presentation

The Numerical Algorithms Group Combining mathematics and technology for enhanced performance Describing and Discovering Mathematical Web and Grid Services Mike Dewar NAG Ltd Mike.Dewar@nag.co.uk Over 30 years of mathematical excellence NAG


slide-1
SLIDE 1

Over 30 years of mathematical excellence The Numerical Algorithms Group Combining mathematics and technology for enhanced performance

Describing and Discovering Mathematical Web and Grid Services

Mike Dewar NAG Ltd Mike.Dewar@nag.co.uk

slide-2
SLIDE 2

Over 30 years of mathematical excellence

NAG Background

Develops and sells software for solving complex

mathematical problems

libraries of numerical algorithms (Fortran, C) visualisation environment (IRIS Explorer) statistical packages (Genstat, GLIM) compilers and tools (Fortran 95)

Undertakes specialised software engineering tasks

maths libraries for Intel and AMD

Customers in many application areas across industry and

academia

Growing business as a component supplier to other

software companies

slide-3
SLIDE 3

Over 30 years of mathematical excellence

NAG and Grid Services

Web/Grid service architecture

Take advantage of new application development paradigm

Service/Software descriptions (MONET)

Semantic web Brokering grid/web services

Focus on capabilities and properties of software rather than its

interface

Portable grid applications for HPC (SciParc)

Platform-independent software layer for grid computing Take advantage of properties of a given architecture including

availability of particular resources and components

Move beyond data grids to algorithm grids

Collaborative Visualisation (gViz Project)

slide-4
SLIDE 4

Over 30 years of mathematical excellence

MONET – Mathematics On the NET

2 Year FP5 RTD activity Principal goals:

Design an architecture for brokering mathematical web

services

Develop the necessary mechanisms for describing

services, problems and queries

Build prototype brokers and services based on generic

technologies

Partners:

NAG Ltd, Stilo Technology Ltd, Universities of Bath,

Eindhoven, Manchester and Western Ontario, I3S Nice

slide-5
SLIDE 5

Over 30 years of mathematical excellence

Simplified MONET Architecture

Service Manager Planning Manager Client Manager BROKER Service Service Client

slide-6
SLIDE 6

Over 30 years of mathematical excellence

MONET Protocols

  • Mathematical Problem consists of

1.

Concrete problem

2.

  • r Problem description
  • Mathematical Query consists of

1.

Problem

2.

Logistical information

  • Mathematical Service Description consists of:

1.

Functional Description

2.

Implementation Description

3.

Service Interface Document

4.

Service Binding Document

5.

Service Metadata

slide-7
SLIDE 7

Over 30 years of mathematical excellence

MSDL: Functional Description of Service

  • Reference to Mathematical

Problem Description Library

  • Input:

1.

F(v1…vn): Rn → R

2.

A ⊆ Rn

3.

{Di : Rn → R, i=1..n }

  • Output:

1.

x ∈ A

2.

f ∈ R

  • Pre-Conditions

1.

Di = ∂F/∂vi

  • Post-Conditions:

1.

F(x) = f

2.

There is no y ∈ A | F(y) < f

Taxonomies

e.g. GAMS G2h1a2 Provide a hook into UDDI

Semantics Supported

OpenMath CDs

Directives

Find

Semantic Description

RDFS/OWL + OpenMath

slide-8
SLIDE 8

Over 30 years of mathematical excellence

The Matching Process

Query and service descriptions are converted to

OWL documents using MONET ontology

Use generic Instance Store component to find

service documents which are consistent with the query

Assemble plans for solving problem using generic

  • rchestration language e.g. BPEL4WS

Return those plans to user with details of how

closely they match

slide-9
SLIDE 9

Over 30 years of mathematical excellence

SciParc

Applications: Quantum Chromodynamics, Biophysics,

Hydrodynamics, …

Software: well-defined components and building blocks for

assembly and deployment of application codes in a grid setting

Flexible and general enough for real applications Capable of efficient implementation on range of hardware Modular, extensible structure rather than rigid API

Hardware: massively parallel systems (apeNEXT,

BlueGene/L, large clusters, …)

slide-10
SLIDE 10

Over 30 years of mathematical excellence

Summary

Very little real demand for commercial grid/web service

software in our customer base (except .NET!)

Existing tools and infrastructure not really industrial strength

WSDL support limited OWL tools incomplete Little in the way of existing ontologies and metadata

However indirect use of MONET-style technology in-house

very promising

specification of algorithms/software

used to embed code in different environments/customise codes for

particular clients

used to customise documentation

can envisage automated assembly of grid applications based on these

specifications

slide-11
SLIDE 11

Over 30 years of mathematical excellence

Further Information

NAG:

http://www.nag.co.uk

MONET:

http://monet.nag.co.uk

SciParc:

http://www-zeuthen.desy.de/sciparc