The MPLS-OPS Archive

Cell Relay Retreat>MPLS-OPS Archive>month:2002-Aug> msg00051



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

Re: UTI (Universal Transport Interface)

  • From: Yakov Rekhter <yakov@juniper.net>
  • Date: Wed, 07 Aug 2002 05:21:21 -0700
  • cc: rcheung@houston.sns.slb.com, mpls-ops@mplsrc.com
  • Resent-Date: Wed, 7 Aug 2002 09:34:33 -0400
  • To: Luc De Ghein <ldeghein@cisco.com>

Luc,

> Raymond,
> 
> > Chris,
> > I can't find any info about "universal VPN" or "L2TPv3" from CCO. I find th
e
> 
> Please have a look at :
> 
> http://www.cisco.com/warp/public/cc/so/neso/vpn/unvpnst/2tpv3_ov.htm
> 
> 
> > UTI info in 12.0S and 12.0ST IOS releases but it doesn't give much
> > background info.
> > Does "AToM" stand for "Any Transport over MPLS" or just for ATM ("ATM over
> 
> "Any Transport over MPLS"
> ATM was the first one we had.
> 
> > MPLS")? Is FR also supported? The doc I got from CCO is always based on ATM
> 
> Please see at the bottom of the page of URL above.
> 
> > only, never mentions a word about other layer2 technologies like FR. The wa
y
> > to do ATM and Ethernet are also quite different. EoMPLS doesn't require you
> > to configure a tunnel but using keyword "l2ptransport" while AToM does. So
> 
> Ths CLI is different for ATMoMPLS vs EoMPLS.
> 
> > that's why I'm guessing AToM doesn't apply to any transport though I heard
> > this since last year.
> > So I guess UTI can generally replace GRE tunnels, what benefits from UTI?
> 
> UTI can transport layer2. GRE transports layer3.

Since one could run MPLS over GRE, and since AToM can transport layer2,
it follows that a combination of GRE+MPLS can transport layer2.

Yakov.

> Thanks,
> 
> Luc
> 
> > well, I think all answers depend on the info I'm looking for.
> > Appreciate if you can find out some info regarding to UTI or L2TPv3. (like
> > what's new in L2TPv3 comparing to v2, etc).
> > 
> > 
> > best regards,
> > Yinglam (Raymond) Cheung
> > Double CCIE #6961
> > Senior Network Engineer
> > Schlumberger Network Infrastructure Solutions
> > Tel: 1 (713) 513-3285
> > Cell:1 (832) 274-0126
> > 
> > -----Original Message-----
> > From: Christopher Lewis [mailto:chrlewis@cisco.com]
> > Sent: Friday, August 02, 2002 12:21 PM
> > To: rcheung@houston.sns.slb.com
> > Cc: MPLS Mailing List
> > Subject: Re: [MPLS-OPS]: UTI (Universal Transport Interface)
> > 
> > 
> > Raymond,
> > 
> > There should be material on Cisco's website under the title of "universal
> > VPNs" which will give the marketing positioning of these two technologies.
> > 
> > I'll give you the standard Cisco answers to these questions.
> > 
> > What IETF draft is UTI based on?
> > UTI is the Cisco proprietary pre-standard implementation of L2TPv3, as such
> > you won't see a draft on it. There are plenty of L2TPv3 drafts for you to
> > find on the IETF website. There is very little difference between UTI and
> > L2TPv3 though.
> > 
> > Why Cisco has all similar technologies on the market?
> > Good question. Cisco wants to remain the number one supplier of networking
> > equipment, as such, it cannot offer just one solution to a problem, as no
> > one solution appeals to all customers.
> > 
> > Does anyone has a list of pros and cons of Cisco's UTI as compared to
> > EoMPLS or AToM? What IETF draft is UTI based on?
> > Well, first EoMPLS is part of AToM. AToM is about transporting any layer 2
> > frame over an MPLS network, as such, ethernet is one of those layer 2
> > technologies transported. Next, UTI/L2TPv3 is for transport of layer 2
> > frames over IP cores, whereas AToM is there for layer 2 transport over MPLS
> > cores. UTI/L2TPv3 provides the same sort of functionality as AToM, but a
> > customer does not have to implement MPLS in their network if they chose not
> > to upgrade software on their core routers (in either case, upgrading the
> > edge routers are of course necessary). UTI/L2TPv3 was created based off
> > customer demand for this functionality.
> > So there is little point in listing the pros and cons, basically the
> > customer will decide which is better depending on whether they want MPLS in
> > their core or not. Currently, the majority of customers have opted to
> > implement MPLS for this functionality.
> > 
> > Hope this helps
> > 
> > Chris
> > 
> > At 07:11 PM 8/1/2002, Yinglam Cheung wrote:
> > >Does anyone has a list of pros and cons of Cisco's UTI as compared to
> > EoMPLS
> > >or AToM? What IETF draft is UTI based on? Why Cisco has all similar
> > >technoloies on the market?
> > >I tried the example on CCO document and it doesn't work (while using EoMPL
S
> > >it works). Does UTI have any debug commands (other than "debug tunnel") to
> > >troubleshoot?
> > >
> > >best regards,
> > >Yinglam (Raymond) Cheung
> > >Double CCIE #6961
> > >Senior Network Engineer
> > >Schlumberger Network Infrastructure Solutions
> > >Tel: 1 (713) 513-3285
> > >Cell:1 (832) 274-0126
> > >
> > >-------
> > >The MPLS-OPS Mailing List
> > >Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> > >Archive: http://www.mplsrc.com/mpls-ops_archive.shtml
> > 
> > -------
> > The MPLS-OPS Mailing List
> > Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> > Archive: http://www.mplsrc.com/mpls-ops_archive.shtml
> > 
> 
> -------
> The MPLS-OPS Mailing List
> Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> Archive: http://www.mplsrc.com/mpls-ops_archive.shtml

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml