perfsonar development work

[Andreas Hanemann <>]

Hi Maurizio,

For the Cambridge meeting I have read the submitted version of
deliverable DJ1.2.4 and would like to send you some comments.

I appreciate the way the three different types of authentication
scenarios are described in section 3. In order to avoid
misunderstandings, I think it would be good to directly match those
scenarios to the clients and tools that we actually have. For
perfSONARUI the document directly says that it is a client acting on
behalf of a user, but I am not completely sure whether VisualperfSONAR
would be the same or the Web Container case. For CNM the situation may
be more difficult. With the current information collected it can be
regarded as automated client since the data retrieval from the CNM and
its data collection are separated. However, there can be different ways
to realize a customization of views using eduGAIN in later stages which
we should discuss in the meeting.

IMHO, it would be preferable if some more information was provided in
section 4 and if the linking of the parts was more obvious. For example,
a bit of overview of the classes in eduGAIN could be provided (e.g. what
is part of the eduGAIN base).  For the client part (4.2) please give an
overview of the steps that have to be carried out by perfSONAR people in
the implementation because the relations of the parts are not clear
otherwise. The same holds for 4.3 and 4.4. For 4.2.2 a bit more
information should be provided. It is not clear whether you strongly
recommend to use the standard implementation or whether perfSONAR
implementors should take a look at different implementations.

In 4.3.1 a reasoning should be given why the PEM format should be used.
As I am not an AA expert and do not know the format which may also hold
for other perfSONAR people, I would appreciate it if a few sentences
were spent on it. 4.4.2 seems to correspond to 4.3.2., but a statement
whether it is covered by eduGAIN is missing.

In sections 4.4.3. and 4.4.4 there are some statements using
"shall","may","could". I am not sure whether the decision on this points
is dependent on the preferences of the implementors only. Do you intend
to make decisions about these points in the Cambridge meeting?

Best regards
Andreas

-- 
Andreas Hanemann, 

Boltzmannstrasse 1, 85748 Garching, Germany
Telefon:   +49 89 35831-8712
Fax:       +49 89 35831-9700