IETF-71 Routing Area Open Meeting
=================================

http://rtg.ietf.org

ADs: Ross Callon, Dave Ward

Scribe: Deborah Brungard 

Administriva and Area Status
----------------------------
 
Ross opened and reviewed the agenda.


NATV4V6 Forwarding Demo Overview (Alain Durand)
------------------------------------------------
Reviewed slides.
Problem Statement:

Several scenarios are being looked at - increasingly complex networks,
as larger networks have very expensive opex costs. Trying to simplify.
Advantageous to use Double NAT v4v6v4. Though need to do v4/v6 translation
and need gateways. This does provide a migration path to v6.
Proposing a simplification for this approach with Tunnel + v4 NAT. This is
a prototype code which we are running, we will be doing also Dublin,
if you are interested to participate, we welcome.

Philip Matthews: I'm interested in what you see as the pros and cons
of the tunnel approach?
Alain: Thinks the tunnel approach is simpler.
Philip: What is the difference that you see between these?
Alain: Difference where you put the application gateways.
Philip: Do you really need an application gateway?
Alain: Yes
Philip: Other differences?
Alain: Some people more comfortable with tunnels, others or not.
Differences of views on how easy it is to debug.



 
Routing Area Working Group Reports
----------------------------------

BFD (Jeffery Haas) - last called two documents. Base document, received
only a few comments. Expect to submit to ADs before Dublin. Also
respinning MIB, a few more IETFs before complete. Starting work on multicast.

CCAMP (Adrian) – Meeting 2x. CCAMP has two meetings this time. Hope
to soon get back to one meeting very soon. 4 RFCs finished.
Two in queue. Several ready for last call (multi-layer, interdomain recovery).

FORCES (Dave W) – didn't meet. Model draft is in. New working group
chair though couldn't attend so didn't meet.
 
IDR (Dave W) – There will be a discussion on Dickerson draft.
BGP TE draft will go to softwires and Lou's document will go there.

ISIS (Dave W) – Almost all documents are done to move from information
to proposed standard. A presentation on multi-instance draft.
Gen-app will be going to last call. BFD bootstrapping is going to last call.
There will be a discussion to remove pseudonodes
from isis. The trill profile work on isis will be done in isis.
 
L1VPN (Adrian) – Four drafts have all completed last call and in
IESG evaluation queue. We've finished all our work so if no new work
comes up, we can close.
 
MANET – No one here.
 
MPLS (George) – Somewhat quiet meeting. Discussion on tmpls, if
option 1 some work will be created. Interest in some oam work.
Security work almost done. Multicast draft in last call. Once it gets
thru other work can start going thru.

OSPF (Acee) – Finishing some drafts, in IESG chats currently.
OSPFv3 respin is going to be done, and some other documents are
getting implemented. MIBs almost done. Updates on some of the
multi-instance drafts. MANET - 3 drafts will go thru as experimental
before Dublin. We will be having a BOF on security key management
(KMART) - discussing current deployment and different reqs for
bootstrapping key management and interactions with routing.
Dave commented - security has held up several of our documents
because we don't have key management, so we need to work with the
security folks, recommend for you to be there so we can discuss,
and be interesting for operators to show up and state their needs.

PCE (JP) – Finally agreed on PCE-IGP needs for that draft, finished
PCEP, almost 10 implementations and several interops (among 4 vendors).
Will discuss this afternoon on rechartering for multipoint support.

ROLL (JP) - Finally got kicked off. We plan to have requirements
done for Dublin.
  
PIM (Stig) – Several documents going on. IPsec for multicast waiting.
Starting some new work, PIM refresh reduction. New work on debugging
for multi-cast.

RPSEC (Tony) – BGP security finished last call. Nothing really left.
So one last look at charter, and consider closing.

RTGWG (John Scudder) – Not meeting. Will meet in Dublin. Fair amount
for that meeting. Last calling again the IPFRR LFA draft, mibs fixed.

SIDR (Sandy) – Energetic meeting, we had Jeff give an overview of how
SIDR will work in operational environment. Reviewed certificates draft
and operator views. Will clarify for drafts that work on a
subset of BGP security (not all scenarios). Will re-address some
previous issues which now have some ways to resolve.
Dave commented - now is a good time for others to start paying attention
as ready now to get more in routing issues.

VRRP (Dave W) – didn't meet.


MEXT Chairs on Routing in Airplanes (Marcelo)
--------------------- 
Reviewed slides.
Requirements from aviation community.
Challenges include security, scalability. Scalability looking for input
from the routing community with expertise on global routing to review
the solutions being proposed to ensure will accomplish the goals.
Chops: trying to optimize round trip time?
Marcelo: yes
Chops: should put up front. I can see concern for some items
but don't see need for all, making more complex than need.
William: for pilot-air control related, do need, and others, consumer,
probably not.
Chops: we should try to keep simple though to. If design for
air traffic control should consider carefully if need such stringent specs.
William: round trip time is critical though for take off and landing.
Chops: I'd rather have a simple solution with one tunnel with 2 sec than
an overdesigned system with more chance to fail.
Marcello: depends how much want to optimize, we don't yet know how
complex the solution will be. 
Chops: doesn't a single one work?
Marcello: it depends, maybe don't need mobility. We are still doing
requirements.
Philip: I don't quite understand requirements. Are you not
communicating directly with the air traffic controller?
Marcello: they say they need mobility. They want to keep communications
while try to change among different technologies.
Lut(?): Couldn't you say for alot of this could do with re-numbering?
Only for items where don't control both ends then could do like proxies?
Marcello: But that would only be valid when in one location,
when I asked them, they say they need mobility.
Lut: Well hope they realize the complexity and the risk of failure.
Marcello: We are evaluating that even if breaks, the protocol still works.
John: Most are commenting similar to my thoughts, the draft should
be made really clear on the assumptions and use cases. Especially
recognizing the complexity of using IP technology. And the costs
involved with enginnering for different reliability.
William: They are trying to move to digital and more security.
Looking for new technologies (using satellite today, low bandwidth),
and low cost (reuse tools available today).
Acee: Is this because the current optimizations RFC3775 are not
sufficient?
Stewart: Actually this is really not a joke about IM - to do
IM for instructions to pilots. Previously had actually looked at
and found chance of pilot error was lower with IM. So use of IM is
always being re-evaluated over the years.
Chunta (?): Why looking at IP?
William: we are looking at other technologies. But need scale and
low cost.
Ross: From this discussion, there is interest in the routing area on this,
even though this wg is in the internet area. We should look at having
the output reviewed here in the routing area wg.
John: We can look at in the routing area wg.
Marcello: ok. As we start looking at solutions, you can help us.

 
T-MPLS Update (Stewart Bryant)
--------------------------------------------------
Reviewed slides.
ITU and IETF will work together to assess option 1 (TMPLS protocol
development in IETF) vs. Option 2 (ITU uses a different ethertype
and develops), Dave and Malcolm Betts (ITU) will lead a Joint Working Team
for the assessment. IETF will also have a MPLS interop Design
Team comprised of chairs and several others to support the JWT.
Already work on enhancements for OAM are being looked at in the MPLS WG.
And a draft will be done to clarify the use of EXP
as there seems to be some confusion on the ITU-side.


Loop-Free IP Fast Reroute Using Local and Remote LFAPs (Ibrahim Hokelek)
      (draft-hokelek-rlfap-01.txt)
------------------------------------------------------------------------ 
John introduced by saying the wg is progressing a method for fast reroute
LFAP. Ibrahim is going to present for information. Ibrahim noted they now
have an implementation.
Stewart - just want to say, we (Cisco) also have a draft which we were
going to present if there was a routing area wg meeting, so we will
present in Dublin.
Ibrahim presented the slides. Noted this is for mobile networks.
John (with wg hat) - your draft is different from other proposals
as they use an equation bounded by the link failure detection time
and local recovery time. Your solution also has the time for signaling
end-to-end. So your solution will take longer than a purely local
one.
Ibrahim - we believe it's just a few ms more.
John - I encourage IGP implementers to speak on this.
Dave - also this will be discuss in Dublin.
Stefano - Other solutions are more deterministic as local (standalone
solution) whereas yours also needs interop and signaling considerations.
Ibrahim - yes but also not so much more.
Stefano - we already have LFA.
Ibrahim - what is the convergence time of yours?
Stefano - within 50ms.