AHM04 1 gViz: Visualization and Computational Steering on the Grid Ken Brodlie, Jason Wood –...
-
Upload
neil-blankenship -
Category
Documents
-
view
215 -
download
1
Transcript of AHM04 1 gViz: Visualization and Computational Steering on the Grid Ken Brodlie, Jason Wood –...
1AHM04
gViz: Visualization and Computational Steering on the Grid
Ken Brodlie, Jason Wood – University of Leeds
David Duce, Musbah Sagar – Oxford Brookes University
2AHM04
gViz – Visualization Middleware for e-Science
• gViz is an e-Science Core Programme project – just finished…
• … has made a start at understanding:
– How to evolve existing visualization systems to the Grid
– How to link visualization and simulation environments
• gViz partners:– Academic: Leeds, Oxford,
Oxford Brookes, CLRC/RAL– Industrial: NAG, IBM UK and
Streamline Computing– International: Caltech, MIT
• Leeds contribution through the White Rose Grid e-Science Centre of Excellence
e-Science Centre of Excellence
3AHM04
Starting Point: Dataflow Visualization Systems
• Visualization represented as pipeline:
– Read in data– Construct a visualization in
terms of geometry– Render geometry as image
• Realised as modular visualization environment
– IRIS Explorer is one example– Visual programming paradigm– Extensible – add your own
modules– Others include IBM Open
Visualization Data Explorer
data visualize render
… pipeline design done after committingto particular system… modules assumed to execute locally
BUT
4AHM04
Extending the Reference Model to Grid Environments
• Revisit the visualization pipeline– Start with the traditional
reference model– Progressively bind in software
and hardware resources– Three-layer reference model
• Conceptual: intent of the visualization
– Show me isosurface of constant temperature
• Logical: bind in the software system
– Use IRIS Explorer (or vtk, or whatever)
• Physical: bind in the resources to be used
– Run the isosurface extraction on particular Grid resource
data visualize render
5AHM04
Developing an XML Language for Conceptual Layer: skML
• First – the conceptual layer
• Dataflow consists fundamentally of:– a map – containing links – between ports – on modules– which have parameters
• This leads us to a simple XML application for visualization: called skML
• Here a data reader is linked to an isosurfacer
<?xml version="1.0"?><skml><map><link> <module name="ReadLat” out-port="Output"> <param name="Filename"> testVol.lat </param> </module> <module id=“iso”
name="IsosurfaceLat" in-port="Input"> <param name="Threshold" min="0" max="27"> 1.8</param> </module></link>…
6AHM04
Diagrammatic Representation using SVG
• skML gives us an XML application for visualization at the conceptual layer
• In addition to language representation, a diagrammatic representation has been created in SVG – so we can do dataflow programming in a web browser
• Transforming to the logical layer binds in the software resource
– A new IRIS Explorer module can read skML and generate corresponding map
– skML can also be turned into an IBM Open Visualization Data Explorer network
<?xml version="1.0"?><skml><map><link> <module name="ReadLat” out-port="Output"> <param name="Filename"> testVol.lat </param> </module> <module id=“iso”
name="IsosurfaceLat" in-port="Input"> <param name="Threshold" min="0" max="27"> 1.8</param> </module></link>…
7AHM04
Physical Layer – Secure Distributed IRIS Explorer
IRIS Explorer on multiple hosts
Select remote host
Automatic authentication using: •Globus certificate
•SSH Key pair
• Moving to the physical layer, we need to be able to execute modules on remote Grid resources
• IRIS Explorer has been extended to allow a user to place modules on specific compute resources – dataflow pipeline thus spans the Grid
• Compute-intensive modules can be placed remotely - design the dataflow for the Grid
8AHM04
Next Steps
• Some tangible benefits…
• … Next release of IRIS Explorer will include the distributed execution facility…
• … but much remains to be done
• Conceptual level– Visualization ontology needed
to define and organize set of canonical processes
– Useful to include resource constraints (initial steps made with RDF)
• Logical level– Visualization data exchange
between systems needs to be studied
– Initial steps made by Julian Gallop (this conference)
• Physical level– User allocation of modules to
resources needs to be replaced by a brokering service
9AHM04
Computational Steering
• Computational steering requires a link between a visualization environment and a simulation environment…
• … gViz library provides this glue
• Design aims:– Use with different simulation
environments and different visualization environments
– Allow connect and disconnect– Lack of intrusion and minimize
performance loss– Robustly handle different producer-
consumer rates– Support multiple simulations– Support collaboration– Support historical audit trail
control visualize
visualization environment
simulation environment
gViz library
gViz library
10AHM04
Environmental Application
• Demonstrator created for an environmental crisis scenario
– Dangerous chemical escapes!– Model dispersion using system
of PDEs and solve numerically over mesh
– Visualize mesh elements where concentration exceeds threshold
– What happens when the wind changes?
– ‘faster-than-real-time’
• Simulation environment– Finite volume code written in C
11AHM04
Pollution Simulation Using the gViz Library and IRIS Explorer
DiscoverGrid resources
Launchsimulation
Connectto simulation
Send controlparameters
Get datato visualize
Visualize
12AHM04
IRIS Explorer as Visualization Environment
• Distributed module execution:– Allows visualization modules to
be collocated with simulation to minimize data traffic to desktop
• Collaborative visualization:– Allows the COVISA multi-user
visualization facility to be exploited
13AHM04
Pollution example with other visualization environments
• Different visualization environments can be connected through gViz library to the underlying simulation
• Note that multiple users – with multiple visualization environments – can connect… allowing collaboration amongst a team
SCIRun
Matlab
vtk
14AHM04
Computational Biology
• In another application the gViz library provides monitoring and control of heart modelling experiments – Arun Holden & Richard Clayton
• Multiple simulations of electrical activity of the heart
15AHM04
gViz Anatomy
DiscoverGrid resources
Launch simulation(register with Directory Service)
Call up Directory Serviceand select simulation
Visualize Visualizemultiplesimulations
Getresults
16AHM04
…Or with Matlab as Visualization Environment
17AHM04
… Or with Grid/Web Services approach
• Grid service interface to gViz library
• Heart Modelling Grid Service uses:
– Web interface where user specifies user name and passphrase, and location of gViz directory service
– Grid service connects to simulations to allow steering parameters to be sent, and results to be retrieved, via the gViz library
– A second grid service builds images from simulation data
• Returned as a Web page
18AHM04
gViz meets Integrative Biology
• The application to heart modelling continues in the Integrative Biology project with David Gavaghan
• Here Matlab is the simulation environment …
• .. linked by gViz library to IRIS Explorer as the visualization environment…
• … or indeed Matlab can act as the visualization environment
• Reality Grid steering also being used in IB project, so hope is to gain convergence between the two approaches
19AHM04
Conclusions
• The gViz project has begun to explore the issues in evolving visualization systems to Grid environments
• Tangible benefits:– Secure distributed IRIS Explorer in next release from NAG– gViz library code will be made available as open source (LGPL)
• Raising issues:– Ontology– Visualization data exchange– Visualization brokering service
• Continuing development of gViz library within Integrative Biology – with potential convergence with RealityGrid steering library
• Demonstration: WRG Stand, Friday 10.30 – 14.30
20AHM04
Acknowledgements
The gViz project team has involved many people:
• Leeds University: Ken Brodlie, Jason Wood, Chris Goodyer, Martin Thompson, Mark Walkley, Haoxiang Wang, Ying Li, James Handley, Arun Holden, Richard Clayton (now Sheffield)
• Oxford Brookes University: David Duce, Musbah Sagar• Oxford University: Mike Giles, David Gavaghan• CLRC/RAL: Julian Gallop• NAG: Steve Hague, Jeremy Walton• Streamline Computing: Mike Rudgyard • IBM UK: Brian Collins, Alan Knox, John Illingworth• CACR, Caltech: Jim Pool, Santiago de Lombeyda, John McCorquadale• MIT: Bob Haimes
Development environment at Leeds: White Rose Grid – e-Science Centre of Excellence