OGSA-DQP is built on top of OGSA-DAI to provide a parallel query capability.
Could be useful for large queries on distributed GI resources.
Requires OGSA-DAI wrappers around all the databases.
Idea is to minimise the amount of time needed to render the view (map).
OGSA-DQP being integrated into OMII grid middleware stack.
2006-03-02
Portals
Rob Allen, Xiaobo Yang (Java portal developer), Xiao Dong Wang (Java portal developer)
Adrian Fish, Miguel Gonzalez-Losa
Portal is a presentation layer that sits inside a browser and hides the complexity
to do with accessing and modifying resources. A user just needs to sign in and
they can do what they can (Role Based Access Control). Offers a means of having a
single sign on to all your compute resources via a URL web page.
MyPROXY server is essential to be able to access Grid Resources from any computer.
Basically this generates a temporary certificate from the credentials of your real
certificate that has a well defined lifetime...
As a user you generate your proxy certificate using myproxy.grid-support.ac.uk
MyPROXY is only essential while individual users must look after their own certificates.
If we develop a portal and want to share it Rob has a registry. Let him know and
others can find it easier...
GI community is doing it all with port 80 using webservers and http. What advantage
do they gain by Grid? More ports more resources er ... that's it!
Provide URL links to Home Pages of attendees where possible.
Make wireless available for all participants.
Make all presentation slides available on line as each session is done.
Evaluation
Should have been doing this as things progressed...
Thoughts
Need to find out more about who everyone else is...
Consider fellowship application for joint NSF ESRC e-Research...
Find out more details from Gillian.
Paul Townend would be interested in the fault tollerance (or lack of it)
in the current version of OGSA-DQP.
Use a bugzilla bug tracking system for MoSeS code development.
For MoSeS plan to get access to copies of all the UK data we can, integrate it
how we can, then go back to all the data providers and ask them if we can go
public with specific end results...
This should gradually draw the data providers/licensors into consultation and
the process of defining what data access levels integrated data have....
To do this we needd to think about producing good metadata i.e. properly
scaled provenance information...