""John Neiberger"" wrote in message ...
> I'm just now digging deeper into current VPN technologies since
> I'm researching Qwest's PRN service. I'm awaiting a definitive
> answer from them but it appears that their PRN service is 2764-based,
> which apparently means it does not use MPLS like 2547-based VPNs.
> I'm curious about the implications of choosing one model over the
> other.
Use the source, John. Read the RFC's, think of the protocol
complexity, check and see if there is any opensource working code
(or even examples of code)...even if you don't understand code,
you understand `wc -l', right? Now, think of the popularity aspects
as well...how many good coders would implement it, and how long
have they had to debug it?
There are other factors to consider...especially when looking at
a vendor. What is the problem you are trying to solve? What are
your requirements for your service? For some reason, I doubt your
VP or Director is going to reward you for implementing something
with "MPLS" in the service offering....but they might reward you
for replacing Frame-Relay or ATM (or even IPSec-based iVPN's)
with a more cost-effective, or better working, solution.
Finally, when deciding on a product you have to look at all the
non-technical requirements as well.......especially cost (both the
hard costs and the soft, hidden costs). Supportability (for you
and for Qwest, and for Qwest's equipment vendor of choice, etc).
There are lots of things to consider, really.
> I thought the market trend was toward MPLS-based VPNs but 2764
> seems to argue against that. What are the implications of choosing
> one model over the other? Are there any major drawbacks to either
> one that the other addresses?
How about AToM vs. MPLS-VPN or 2547bis vs. Martini? Or Martini
vs. Kompella? Or Kompella vs. Kompella (do a LR search for that,
it's really funny)? What about L2TPv3/PWE3 vs. any of the above?
Does one consider GRE or IP-in-IP? IPSec? How about Interworking
(also known as mix-and-match)?
When you say MPLS-VPN, I immediately think 2547....however, that's
becoming less-and-less true. In fact, I don't believe the two
largest successful offerings of MPLS-VPN (Level-3's (3)Packet and
GX's Smart/ExpressRoute) use 2547 MPLS-VPN's. They use something
else. But other vendors don't even use MPLS-VPN to solve the CE-VPN
(or even IP-VPN) problem.
> I'm also a little concerned about vendor choices. Nortel seems to
> be pushing 2764, while Cisco and possibly Juniper are pushing 2547
> and MPLS. Is that correct? If so, is that really that important to
> the customer?
Cisco is/was pushing two things: UTI and EoMPLS (and now the full
line of AToM), mostly non-MPLS PWE3 or MPLS-VPN with Martini.
Juniper is/was pushing CCC and K.Kompella. Laurel is/was pushing
Martini... this list could go on forever. The new game in town
is TiMetra (who was purchased by Alcatel), with VPLS by V.Kompella.
Everybody now wants VPLS. It's super bleeding-edge, but the
technology works great (on paper).
> Forgive me if these questions seem pretty vague. I'm still learning
> about the technologies involved and I'm not very familiar with the
> specifics and the terminology.
You should read LR (Lightreading) and go to SuperComm if you really
want to "get into" this stuff. If you really want to understand
it, well then you only have one option: RFCs, loose consensus and
running code.
I, personally, do not want to get heavily into it. It's not mature
technology, and it's all bad, IMO. There are a few solid
technologies...and they are mostly the ones that were implemented
first. Sure, MPLS-VPN with 2547 is great, but it scales horribly
and is difficult to manage. It also is dependent on an MPLS core,
with no route summarization, full IBGP with MP-BGP, and all the
rest of the hooks. It uses two (and if you are using MPLS-TE, then
three) labels to work, so it's big and kludgy. And you add all
sorts of bugs and overhead to get it working.
So if MPLS-VPN with 2547 is poor, think about how all the rest of
this stuff also breaks things all the time. It's all poorly implemented!
If you want it, you have to pay the price for it. IMO, I think
UTI and L2TPv3 with Sprint or C&W have been incredibly as successful
as Level-3 and GX's MPLS-VPN products.......maybe not in Europe,
but that doesn't mean the technology doesn't work there....it's
just that MPLS-VPN is very popular in Europe (and Asia). Don't
make decisions based on popularity contests or superior technology
or even by using what's bleeding or cutting edge. Make the best
decision you can with the criteria you set forward with.
*Don't* follow a market trend.... ;>
> I'll put in a plug here for Howard's book _Building Service Provider
> Networks_. Among a number of things it discusses some of these VPN
> technologies and has been very helpful the last couple of days
> during my research.
Here's some good reading that I recommend (Howard's book is also
vv good!). Warning: it's quite a lot of [EMAIL PROTECTED]:
bash-2.05$ rfc -r
label\|mpls\|atom\|l3vpn\|pwe\|l2tp\|vpls\|2547\|2764\|kompella\|martini |
grep ^[0123] | wc -l
48
bash-2.05$ egrep -i
label\|mpls\|atom\|l3vpn\|pwe\|l2tp\|vpls\|2547\|2764\|kompella\|martini
idlist | wc -l
644
-dre
Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=73050&t=73048
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
