Hey Michael Thanks for you reply to my post. I think the lack of replies is interesting. Quite telling, or maybe I am reading too much into it. ;-) Rgds Peter Michael Newbery wrote:

On 29/11/06 12:35 PM, "Peter Youngquest" wrote:

...

Hey All

I am personally bored with the IPV4-6 discussion, but while everyone is up and talking.

What is everyones opinion in regard to PBT. Does this have a place, can it work happily as a feeder into an MPLS cloud or will it remove the need for MPLS.

Aka IEEE 802.1ah (draft)

<opinion> Tool = hammer => problem => nail.

If you like circuits and never felt comfortable with this newfangled packet thing, then 1ah is a way to make Ethernet feel more familiar to you. Note that it also requires 802.1ag (draft) and QinQ to work, which is not immediately obvious from a first reading. My issue is that it has the Someone Else's Problem property. Actually setting up and tearing down circuits is an exercise left to the reader, whereas it's very much at the heart of what MPLS does.

So, yes, it can work happily as a feeder into an MPLS cloud. If you don't already have an MPLS cloud it might be an attractive alternative. If you do already have an MPLS cloud I'm not sure that you would want to replace it (others may have different opinions).

802.1ah is somewhat of a behemoth: the PBT function is one of 3 things that it does. I *do* think it's useful. From my limited exposure thus far, I don't think it obsoletes MPLS. </opinion>

It is further development on the MAC in MAC metro Ethernet idea, and continues on from PBB (802.1ad and 802.1ah) It gets rid of some of the Ethernet features such as broadcast and links the L2 Shortest path algorithm to ISIS. It is essentially for use in Large metro environments and some have gone as far to say replace or side by side with MPLS. It also hides the MAC addresses from the access/aggregation network through the metro or aggregation to aggregation environment. Say 1k subs = 5k macs are now reduced to 1 chassis mac for the pipe. Having this reduction then reduces the need for learning and obvious reduces the tax on the transit devices tables. There is also discussion that it will allow tighter SLA than MPLS. This is open to discussion esp with all the work going on in the MPLS world. It is being championed by a number of switch vendors and observed by the rest. Time will tell if it gets the acceptance it needs to become mainstream. BT are making noise about, supposedly a Chinese carrier have adopted it (name eludes me). Also a number of carriers (US, Europe and Asia) are keeping it in the watch list and options open during RFP periods. Obviously this is a very quick and dirty overview of what it is. Peter Gerard Creamer wrote:

On 30/11/2006 7:22 a.m., Peter Youngquest wrote:

...

I think the lack of replies is interesting. Quite telling, or maybe I am reading too much into it. ;-)

I'm still reading wikipedia trying to figure out what it does :^) Hard to be opinionated when your clueless. Actually, it's probably easier, but you get flamed...

So if you're reading that you mentioned something outside of the majority's experience, my vote is a 'yea'. Gerard