Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Motion - WAN Attachment

My sentiment on including WAN attachment in the objectives of the first 10G PAR
is similar to that of Tom Dineen's.  This is to say that while it's worthwhile
to consider Ethernet for WAN access or WAN applications, let's not include them
in the goals for the 1st 10GE standard. To ensure continued success of Ethernet
based technologies, products, and 802.3's success, we should keep the objectives
for the 1st 10GE standard to the application spaces that Ethernet
technologies/products/paradigm have proven successful.

From all of the discussions on additional needs as part of applying Ethernet to
the WAN space (i.e. CFI, reflector, interim), it's pretty clear that for
Ethernet-based products to be successful in that space, either the WAN
deployment/operating paradigm has to change or Ethernet standard will have to
change, or may be both, substantially.  Either way, it will likely take a lot
longer to sort things out and to achieve a solid standard with broad industry
(in this case, the carriers) support.  Also, let's remember that WAN attachments
(or WAN access) today are general done with IP routers seperating the LAN and
the WAN, not L2 bridges.  Hence, if 802.3 decides to support the push of
Ethernet into WAN space, it makes the most sense to consider the WAN access and
backbone applications in a seperate PAR in order to do it correctly, kind of as
Jonathen suggested.