Re: IMPORTANT: Strategy for VPN work in IETF

All,

many of the points brought up by Thomas, Jeremy, Ananth and Luyuan
valid.

I would agree that if there is like 13 docs that needs to go into IESG
review shortly for the L3VPNs only, this in itself is proof of that
there is an abundance of work to be undertaken in the L3VPN area.
Haven't checked but if there are MIB modules among those docs, one
shouldn't underestimate the amont of work that needs to go into
preparing those docs for IESG review.

The amount of work is even larger for the L2VPNs.

There seems to a, "if only thinking" around the split. "If only we'd
done this along time ago!" or "It would have been better if we only had
done this from start!".
While I can agree on the sentiment, it is focusing on the wrong problem.

We have a situation where we need to speed up the output in this ppvpn
area, I don't think there is an agreement on this?

My experience from the mpls wg is that I as a wg chair very often
becomes a bottleneck, solely on reeading and crosschecking capacity. Now
if Alex is proposing to throw more manpower on those tasks in an area
that is understood to be work intensive, he nees support in this!

I really think there are separate issues here, but where splitting the
wg could contribute solve more than one, simple by adding fresh blood.

- current work load, more people more work
- generic co-ordination, an important task for wg chairs and the
   tasks would be much clearer
- "wrapping" up of L3VPN, maybe another mindset than in the inventive
   phase
- new charters, more clearly defining where to go
- co-ordination wioth other standards bodies, which is possibly a more
   or less full time task in itself.
- moving the ppvpn work to another area, the current proposal addresses
   my concerns
- closing down the Sub-IP area, don't feel enthusiastic about, but the
   arguments are strong

Conclusion is that if we want to "split work" we do this by sllitting
the wg, and we don't wait, because waiting involves delaying. I won't
claim that the two months to Vienna is "critical", but holding off
decision until after a dual session and starting that process then
probably takes close to the autumn meeting. And going on in a
non-decision situation would be worse than anything else.

I'm ready to say - "Let's jump!" but could understand if a shorter
period of discussion is still needed.


/Loa

Fang, Luyuan, ALABS wrote:
> I agree with Jeremy and Ananth. 
> Luyuan 
> 	-----Original Message-----
> 	From: Nagarajan, Ananth [NTWK SVCS] [mailto:[email protected]]
> 	Sent: Friday, May 09, 2003 11:52 AM
> 	To: [email protected]; Thomas Narten
> 	Cc: Alex Zinin; [email protected]
> 	Subject: RE: IMPORTANT: Strategy for VPN work in IETF
> 	
> 	I think I agree with Jeremy here. After thinking about the proposed split for some time, I realized that (as I said earlier) it would have made more sense if this was done as soon as L2VPN/PWE3-related stuff was brought into the WG. I also feel that there are several "generic" documents, in addition to the ones I brought up on this thread some time ago, and it will become increasingly complex to decide where those kind of documents would fit.
> 	My preference is to schedule two sessions at Vienna, and spend some time discussing this, before any decision is made. 
> 	PS: Another potential problem that would happen is having to subscribe to both the l2 and l3 vpn mailing lists and getting duplicate emails everytime a generic issue is discussed.
> 	Ananth 
> 	> -----Original Message----- 
> 	> From: [email protected] [<mailto:[email protected]>] 
> 	> Sent: Friday, May 09, 2003 9:54 AM 
> 	> To: Thomas Narten 
> 	> Cc: Alex Zinin; [email protected] 
> 	> Subject: Re: IMPORTANT: Strategy for VPN work in IETF 
> 	> 
> 	> 
> 	> Hi Thomas, all, 
> 	> 
> 	> > The chairs themselves seem to agree that there are some 13 documents 
> 	> > on L3 that are not through the system yet. I'm a bit 
> 	> concerned to hear 
> 	> > that there has been minimal discussion on L3 topics over the last 
> 	> > year, given the number of L3 work items that apparently need to get 
> 	> > completed. 
> 	> > 
> 	> > The above could be read as indicating that the discussion/urgency of 
> 	> > the L2 work is pushing out or drowning out discussion on L3 
> 	> topics. If 
> 	> > that is the case, a separate WG, mailing list, etc., may well help. 
> 	> > I'm assuming the L3 work is important and should get completed. 
> 	> 
> 	> I believe that it can also be read as indicating that many feel that 
> 	> most of the (current) L3VPN work is stable and agreed upon, 
> 	> and doesn't 
> 	> need more discussion within the WG. As was already hinted previously, 
> 	> the progress of the L3VPN work has been delayed (for a long time) 
> 	> because the framework and requirements documents needed to be accepted 
> 	> first. 
> 	> 
> 	> I do believe however that several proposed new working items (QoS, 
> 	> security, management, multicast, etc.) have received little attention 
> 	> from the WG because of the current L2VPN (VPLS) 
> 	> discussions/urgency. But 
> 	> as many of these are aimed at both L2 and L3VPNs, and as I don't find 
> 	> reference to these items in the proposed charters, I 
> 	> understand that the 
> 	> goal of the proposed split is not to accomodate these. 
> 	> 
> 	> On one hand, I believe that the idea of two separate WGs is a 
> 	> good one, 
> 	> but on the other hand, I'm not sure how much it will help 
> 	> now, so far in 
> 	> the process. 
> 	> 
> 	> I also personally believe that the PPVPN WG has produced valuable 
> 	> results, and is making progress on complex topics every day 
> 	> (look at the 
> 	> current IEEE vs VPLS discussions on this list). And this despite the 
> 	> fact that the 'market' is not waiting and despite the fact that the 
> 	> number of interested individuals (from different backgrounds: SPs, 
> 	> vendors, academic and other standard bodies) is very large. 
> 	> 
> 	> Jeremy. 
> 	> 
> 	> 
> 	> 
> 
> 
> 
> 
> 


-- 
/Loa

mobile + 46 739 81 21 64
email: [email protected]