RE: Testbed

["Gregor von Laszewski" <gregor@mcs.anl.gov>]

GRAM:

	actually it is both. A protocol and an API (Globus provides a CAPI).
Various CoGs provide API's in Java, Python (through swig). The Java client
API is 100%. I will ask the Globus web page maintainer to put the
specification of the GRAM protocol on the Web.

Gregor


-----Original Message-----
From: Chip Watson [mailto:chip.watson@jlab.org]
Sent: Wednesday, June 13, 2001 4:20 PM
To: Ian Foster
Cc: Gregor von Laszewski; gce-wg@gridforum.org; itf@mcs.anl.gov
Subject: Re: Testbed


Ian wrote:

> ...  Basically, GRAM (say) is an interoperability protocol.

Or is it a library which encapsulates a protocol?  Is the wire protocol
defined
(web link to definition?), and is it easy to integrate into another system?

> To put things another way, Chip says "a homogeneous Globus GCE testbed
does
> not test anything which is not already known to work" [and by definition
is
> presumably not useful].

Not at all. I'm acknowledging its usefulness, but believe that there is
functionality
which Globus does not address.

> I think the whole point is that by adopting a
> homogeneous Globus GCE testbed, we can ignore a [relatively small and
> mundane] set of problems that have been solved, and go on to work on other
> more interesting and critical problems: like how to build schedulers, or
do
> data replication, or how we represent policy, etc., etc.

Agree to second half, but perhaps put some things into that scope which you
would not.  In particular, I think web services are part of our future,
leveraging
the huge investment being made there by Microsoft, IBM, Sun and many others.
Should grid forum do something analogous to ebXML?  I find the possibility
interesting...

--
Chip Watson
High Performance Computing Group
Thomas Jefferson National Accelerator Facility
Tel: (757) 269-7101
http://www.jlab.org/~watson