Venues 2.0 Client

Rick Stevens stevens at mcs.anl.gov
Thu Oct 24 17:53:15 CDT 2002


Some thoughts on Node management.

We've been thinking for some time that we want to broaden the definition of 
a node for the AG to include a variety
of display and interaction devices associated with the core AG 
resources.  An example of this would be a micromural
that is associated with a node.  When the micromural is to be used with the 
AG node it would need to be registered
somehow with the node as an available capability and we would want to be 
able to discover its presence etc.   Seems
that the AG node resource collection would need to be managed as a 
collection of resources (have some description
of the associated devices etc.), have some ability to manage those resource 
(reserve, allocate, digitally control,
install and launch programs,  check status etc.) and to 
monitor.   Informally I've been thinking about nodes
at two levels.  Those that are part of the installed infrastructure (likely 
to be part of the dedicated systems
in the room, etc.) and those that are temporarily associated with the node 
(laptops, pdas etc.).  Those in the second
category might want the ability to associate themselves with the main AG 
node such that venue transitions are
automatically propagated etc.   Its not clear to me currently how to think 
about security for the node as a whole.
Security models for the Grid seem to be based on the concept of a user, but 
the AG node is more likely to be associated
with a space or group, even though their might be an operator as well.  I 
dont know if we have talked to the Globus
folks about Group certificates.  But that might something for us to do.

In the back of my mind I'm thinking that we need some kind of active space 
services that tie things together in a way
that would enable one to write (desgin?) applications for entire rooms.

At 12:26 PM 10/18/2002 -0500, Thomas Uram wrote:

>I've been working on implementation of the Node Management infrastructure, 
>and have been developing a client to exercise the infrastructure.  This 
>has given me a good sense of what the Node Management client needs to do.
>
>We should be able to have a productive discussion, with the different 
>ideas and prototypes we have.  We should meet one day next week when Ivan 
>and Bob are back.
>
>
>Tom
>
>
>Rick Stevens wrote:
>
>>Good idea to collect these requirements.. so at some point we can make the
>>decision if we re implement from scratch or if we can reuse any of the 
>>pre-prototype.
>>
>>At 11:24 AM 10/18/2002 -0500, Terry Disz wrote:
>>
>>>Ivan,
>>>
>>>The venue client we are now using is what I call a pre-prototype. I am sure
>>>Bob can elaborate when he gets back, but from my point of view, it serves
>>>two purposes: To exercise the VV functionality and to provide a testbed for
>>>the docking client. Towards that end, it is more or less hard coded to a
>>>venue but does exercise the venue interface for enter, (but no navigation),
>>>it implements all security Bob has been working on, it exercises the venue
>>>interface for user objects, it has no node management functions, it provides
>>>a client venue interface so the docking tools can gain access to user and
>>>venue information. These latter functions were all in the original WS
>>>docking architecture. I continue to expand it as needed for technical
>>>research issues or other experiments. You can find it and the entire code
>>>base in the fl cvs tree under VV2. I am running an experimental version now
>>>that I have not checked in, which has added functionality.
>>>
>>>All that being said, it certainly is high time we started to figure out what
>>>the entire suite of requirements are, including UI, for a venues client we
>>>want to deliver. I can start by extracting requirements of the Venues Client
>>>from the docking documents and maybe Tom or you can contribute some node
>>>management requirements.
>>>
>>>Terry
>>>
>>>
>>>
>>> > -----Original Message-----
>>> > From: owner-ag-dev at mcs.anl.gov [mailto:owner-ag-dev at mcs.anl.gov]On
>>> > Behalf Of Ivan R. Judson
>>> > Sent: Thursday, October 17, 2002 2:16 PM
>>> > To: 'Ag-Dev at Mcs. Anl. Gov'
>>> > Subject: Venues 2.0 Client
>>> >
>>> >
>>> >
>>> >
>>> > Howdy,
>>> >
>>> > I'd like to start a discussion about a key part of AG 2.0 that isn't
>>> > represented yet by any of our arch or design documents. The Venues
>>> > Client.
>>> >
>>> > I know Chris has been working on some form of a client, but I'm
>>> > wondering what discussion has taken place that's driving that client
>>> > development. Can someone summarize the current client development, the
>>> > intent of the client, and point me at any design documents that might
>>> > exist for the client? That'd be a great place to start this discussion.
>>> >
>>> > Thanks,
>>> >
>>> > --Ivan
>>> >
>>> > ..........
>>> > Ivan R. Judson .~. http://www.mcs.anl.gov/~judson
>>> > Futures Laboratory .~.  630 252 0920
>>> > Argonne National Laboratory .~. 630 252 6424 Fax
>>> >
>>> >
>>
>>
>
>




More information about the ag-dev mailing list