Gamze,
> 1. With regards the email below, which I found from the archive,
> has there been any discussion (or are there any recent results) on
> the performance evaluation of an edge module w and w/o ICAP
> deployment?
No, not that I'm aware of. We did some measurements in our Lab, but
the results were just as expected and not that interesting, so we
didn't publish anything about it. In short, we measured the delay
introduced by the protocol engine and didn't find much of an overhead
for processing the iCAP messages themselves - if I recall correctly,
time/processing capacity required for rule processing and for service
execution, in particular, was much more a factor.
> 2. Is it safe to say that a
> L7 switch (web switch, e.g. nortel alteon) is a simplified OPES
> device (since ICAP is not an requirement)? If not, why?
See comments from Ian - a "traditional" L7 switch as known today is
very likely to miss important features as required by OPES.
> Minor typos & comments wrt the latest drafts:
Thanks for catching those, the respective editors will surely consider
these in the upcoming versions of the drafts.
Thanks,
Markus