SSC Workshop "Minutes"

The primary purpose of this workshop was to understand the borders
between the various SSCs--both vertical (scientific) and horizontal
(generic support).  It was also to understand the work programs of the
core EGI project, application porting support project, and a
scientific gateway project to avoid duplicated tasks and effort within
the scientific SSC work plans. 

These "minutes" are intended to be extremely short and only to
indicate the main areas where issues may arise.  For details on each
project, please see the presentations attached to the workshop agenda.

Following the presentations, the following topics (in no particular
order) were identified as areas needing more discussion.  Not all of
the points were actually discussed before the session ended.  The
descriptions contain only the major points that were brought up during
each discussion.

 * Sustainability of tasks/Differing project durations

The EC has made it clear that the core EGI operations will not receive
additional funding before FP8.  This means that the money for EGI
operations available in FP7 (25 MEuro) either must be used for a four
year project.  (Or that EGI becomes entirely self-sustaining after the
money runs out.)  There is no equivalent problem for the user
activities although assuming that there will be future user-oriented
calls that are pertinent for the SSCs is a risk.

In general the differing lengths of the projects will make the
interactions between the projects more difficult.  The various SSC
projects will discuss if expanding their work plans to four years to
match EGI makes sense.  However, given the tight funding scenarios it
is unlikely that this will happen.  In this case, each SSC must
seriously consider how to sustain their activities past the end of the
currently proposed projects even without additional EC funding. 

 * New Users/Communities and Expanding existing user communities

It was pointed out that addition of new users and communities can have
a negative impact on sustainability as more resources will be required
for support, maintenance, etc.  However, the infrastructure must to
some extent remain open to new users and new communities and provide
services aimed towards them.

Dissemination and training are critical for expanding the grid user
community.  CUE will provide these services for new users and new
communities; however, this must be supplemented in parallel with
similar efforts within the existing scientific clusters.  The results
of these efforts are often not seen immediately but over time.

 * Requirements gathering and general communication flow

Clearly the requirements gathering process cannot be done in a
scalable manner if all of the SSCs talk with all of the middleware
providers.  EGI has foreseen a middleware coordination board to act as
a central point for facilitate the transfer of information related to
software services.  It is also important that these mechanisms be open
and transparent so that all of the stakeholders can understand how
priorities are set and the roadmap is developed.  However for detailed
technical discussions, it is expected that particular users will work
closely with the middleware developers for certain new functionality.

 * Common concerns (data mgt., usability, ...)

There are certain concerns such as data management and usability that
appear in many of the SSC's work plans.  These concerns are general in
nature and it would probably be better if these were coordinated at
the EGI level, allowing evolution of the core middleware rather than
the development of domain-specific solutions.  

 * Possible overlaps: Scientific Gateways and App. Porting Support

Almost all of the scientific SSCs contain tasks related to deploying
and maintaining scientific gateways and providing application porting
support.  These are two areas where it is critical that the overlaps
are ironed out to ensure that we take advantange of the resources
available.  The action to compile a list of "concerns" for each
project should help clarify the borders between projects in these
areas. 

 * Policies/Procedures/Standards

Not significantly discussed. 

 * Planning relations to other proposals/risks on unfunded projects 

Not significantly discussed.  Although the projects were encouraged to
make reasonable work plans based on the success of most projects, but
also including risks (changes to the work plan) should some of those
projects not be funded.  

To advance, Cal Loomis will ask each of the project coordinators to
write up a short description of where the overlaps concern each
project. These can then be used to clarify those overlaps and allow us
to better prepare for the complex landscape of the EGI era.