[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
notes from 6/24 meeting
- To: tewg-dt@ops.ietf.org
- Subject: notes from 6/24 meeting
- From: Jim Boyle <jimpb@nc.rr.com>
- Date: Wed, 27 Jun 2001 13:55:42 -0400
- Delivery-date: Wed, 27 Jun 2001 11:00:04 -0700
- Envelope-to: tewg-dt-data@psg.com
- User-Agent: Mozilla/5.0 (Windows; U; Win 9x 4.90; en-US; m18) Gecko/20010131 Netscape6/6.01
Lots of stuff missed in these notes, but it hits some high points. Also
these are defintiley not exact quotes, and indeed the general points are
highly likely to be misrepresnted as to who said what - sorry :)
started off on questions...
Tom Reddington: worried about the scope of the questions, and what we
are working to cover (and not)
Dave McDysan: we do need to make some progress on, but there are
definitely some issues, such as definitions of terms, maybe starting
with Wai Lai's draft.
Wai Lai: the current draft is just a starting point for discussion.
Jim: We ought to answer the question list we have, tear into some
conversation on the DT-list,
Tom Reddington: On the questions, in some sense, they get into the
requirements for the design, but there is not much delving into trying
to find out where the faults.
Tim: Yes, that echo's his concerns. His lab likes to break what makes
communications work. Issues like scripts against BGP.
Dave: So everyone should get their answers in on the questions by
Monday or so, and let's kick up discussion on the list, and we'll have
another call.
Jim: Definitely, we should kick up the volume on the list, so does
anyone besides rob have any big need for the hierarchy stuff?
Dave: It seems like one context that might be a definite driver is that
of edge to edge VPNs across potentially traffic engineered data
networks. There area also is a lot of emergence of protocols addressing
different VPN approachs.
Jim: (after some discussion), the VPN application probably is a good
reason to actually be rethinking hierarchy and scalability. It seems
like there are probably two types of situations of concern:
1) large flat routing domains where edge to edge signalling as implemented
today would probably not scale.
2) networks which would like to signal edge to edge, and might even scale
in a limited application, however they are hierarchically routed (e.g.
OSPF areas) and current implementations, and potentially standards
prevent signalling across areas.
Dave: So we'll talk again next thursday at 1pm