The MPLS-OPS Archive[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index][Thread Index][Author Index][Subject Index] RE: MPLS on Foundry and Riverstone
Just because I love condescending attitudes so much :p.. From some random vendor's documentation. .chance The "draft-martini" documents specify a technique by which a layer two Protocol Data Unit (PDU) (such as ethernet, ATM, and Frame Relay) may be tunneled through a MPLS enabled network. This allows for the provisioning of multiple discrete customer networks over the Service Provider backbone, creating in-essence a Virtual Private Network (VPN) for the customer. While it is common to refer to the implementation proposed by Luca Martini et al as "draft-martini" it should be understood that "draft-martini" is actually comprised of two separate documents - specifically draft-martini-l2circuit-trans-mpls-07.txt and draft-martini-l2circuit-encap-mpls-03.txt, both of which currently exist as a work in progress. * draft-martini-l2circuit-trans-mpls-07.txt introduces the concept of carrying layer 2 protocols though an ISP backbone network by prepending an MPLS label to the PDU. The document also describes the introduction of a new Type/Length/Value (TLV) label that is signaled between the endpoints to enable the layer 2 VPN, and the distribution procedure for this new label. * draft-martini-l2circuit-encap-mpls-03.txt specifies the encapsulation details for the various layer 2 protocols to be transported (ethernet, ATM, Frame Relay, etc.). When creating a layer 2 VPN using the methodologies of "draft-martini" it is assumed that an MPLS Label Switched Path (LSP) has been created between two Label Switch Routers (LSR) , using a standard label distribution protocol such as RSVP/TE or LDP. This LSP is referred to as a "tunnel LSP", and serves to carry a number of emulated virtual circuits (VC). These VCs are signaled and maintained between endpoints using LDP and the new TLV as described in draft-martini-l2circuit-trans-mpls-07.txt, or via static mappings manually configured on each endpoint. The emulated virtual circuit can be referred to as a virtual private network as each VC is unique from other VCs within the same tunnel by the use of a "demultiplexer field" carried in each VC's label. This is akin to the concept of the DLCI in Frame Relay or the VC ID in ATM. The VC label will exist at the bottom of a packet's label stack, while the tunnel label (if present) must exist directly above the VC label. This allows the penultimate router to pop the tunnel label and forward the packet to the VPN's egress LSR. The egress LSR will then make the determination as to proper handling of the packet as well as the underlying layer 2 PDU. It should be clearly understood that "draft-martini" does not make provisions for the concept of a multipoint VC. The martini drafts focus solely on point-to-point connections. This is due to the significant technical issues associated with the concept of multipoint-to-multipoint and point-to-multipoint circuits when dealing with layer 2 protocols - especially those of a broadcast nature such as ethernet. A series of IETF drafts are being developed to address this issue, and will be available for comment sometime in the near future. > -----Original Message----- > From: Christopher Lewis [mailto:chrlewis@cisco.com] > Sent: Monday, October 29, 2001 2:54 PM > To: Chance Whaley > Cc: 'Harish D V'; Duane.Wright@186k.co.uk; ktho@cabletv.com.hk; > mpls-ops@mplsrc.com > Subject: RE: MPLS on Foundry and Riverstone > > > When having this discussion, please be clear about which > Martini draft > being discussed, there are two, one for encapsulation and one > for transport. > > Juniper's Kompella draft now specifies martini encapsulation, but not > Martini transport procedures, I am told Martini encapsulation > will appear > in 5.1 of Junos for CCC, although I'm obviously not a Juniper > authority :-) > > Chris > > At 02:13 PM 10/29/2001, Chance Whaley wrote: > > >I believe that everyone that has talked about Juniper and > martini thus far > >in this little thread has referred to them as a transit for > tunnel LSPs, not > >as edge devices. > > > > > > Even over RSVP tunnels (running through a Juniper in > this case).. > > > > > > > > net-iron-400#sh mpls rsvp ses de > > > > > > > >.chance > > > > > -----Original Message----- > > > From: Harish D V [mailto:hdv@peakxv.net] > > > Sent: Monday, October 29, 2001 11:06 AM > > > To: Chance Whaley; Duane.Wright@186k.co.uk; ktho@cabletv.com.hk > > > Cc: mpls-ops@mplsrc.com > > > Subject: Re: MPLS on Foundry and Riverstone > > > > > > > > > Hi guys, > > > > > > I have been watching this string and was surprised to see > that Juniper > > > supports draft-martini. becoz, I have a confirmed info from > > > Juniper itself > > > saying they don't support martini now!. > > > > > > Regards, > > > > > > Harish D V > > > Peak XV Networks > > > San Ramon > > > CA - 94583 > > > > > > 925-242-7459 office > > > 925-381-9967 cell > > > > > > www.peakxv.net > > > > > > "Out of Clutter, find Simplicity. From Discord, find harmony. > > > In the middle > > > of difficulty, lies opportunity." > > > - Albert Einstein > > > > > > > > > > > > ----- Original Message ----- > > > From: "Chance Whaley" <chance@dreamscope.com> > > > To: <Duane.Wright@186k.co.uk>; <ktho@cabletv.com.hk> > > > Cc: <mpls-ops@mplsrc.com> > > > Sent: Saturday, October 27, 2001 8:32 AM > > > Subject: RE: MPLS on Foundry and Riverstone > > > > > > > > > > Hmmm.. Looks like they run martini to me. > > > > > > > > net-iron-400#sh mpls ldp session > > > > Peer LDP Ident: 45.221.0.40:0, Local LDP Ident: 45.221.0.37:1 > > > > Active: no, State: Operational > > > > TCP connection: 45.221.0.37:646--45.221.0.40:11046, > > > State: ESTABLISHED > > > > Addresses bound to peer LDP Ident: > > > > 45.221.0.40 > > > > 127.0.0.12 > > > > 45.220.254.44 > > > > 45.221.40.40 > > > > > > > > net-iron-400#sh mpls vll detail > > > > VLL FDRY_to_Cisco_7600 VC-ID 3740 > > > > State: UP > > > > Vll-peer: 45.221.0.40 End-point: > > > untagged e 1/4 > > > > Local label: 800000 Remote label: 45 > > > > Local group-id: 0 > > > > Tunnel LSP: FDRY_to_Cisco_7600 (tnl0) > > > > > > > > > > > > Even over RSVP tunnels (running through a Juniper in > this case).. > > > > > > > > net-iron-400#sh mpls rsvp ses de > > > > > > > > Ingress RSVP: 1 session(s) > > > > To From State Style Labelin > > > Labelout LSPname > > > > 45.221.0.38 45.221.0.37 Up FF - 103125 > > > > Foundry_to_Laurel > > > > Time left in seconds (PATH refresh: 1 RESV refresh: 26) > > > > Tspec: peak 0 kbps rate 0 kbps size 0 bytes m 20 M 65535 > > > > PATH sentto: 45.221.37.10 (p3/1 ) > > > > RESV rcvfrom: 45.221.37.10 (p3/1 ) > > > > > > > > Egress RSVP: 1 session(s) > > > > To From State Style Labelin > > > Labelout LSPname > > > > 45.221.0.37 45.221.0.38 Up FF - - > > > > LaurelToFoundry37 > > > > Time left in seconds (PATH refresh: 22 RESV refresh: 15) > > > > Tspec: peak 0 kbps rate 0 kbps size 0 bytes m 20 M 4458 > > > > PATH rcvfrom: 45.221.37.10 (p3/1 ) > > > > > > > > > > > > net-iron-400#sh mpls vll de > > > > VLL Foundry_to_Laurel VC-ID 3738 > > > > State: UP > > > > Vll-peer: 45.221.0.38 End-point: > > > untagged e 1/3 > > > > Local label: 800025 Remote label: 103644 > > > > Local group-id: 0 > > > > Tunnel LSP: Foundry_to_Laurel (tnl0) > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -----Original Message----- > > > > > From: Duane.Wright@186k.co.uk [mailto:Duane.Wright@186k.co.uk] > > > > > Sent: Wednesday, October 24, 2001 4:04 AM > > > > > To: ktho@cabletv.com.hk > > > > > Cc: mpls-ops@mplsrc.com > > > > > Subject: Re: MPLS on Foundry and Riverstone > > > > > > > > > > > > > > > > > > > > The Foundry NetIron at present doesn't run MPLS martini > > > > > draft or Kompella > > > > > but there is interop info available. > > > > > Riverstone has already implemented Martini draft and it has > > > > > been tested > > > > > with Juniper's & Cisco. > > > > > As regards the performance if the Riverstone it works well. > > > > > In addition there is another emerging standard the Lasserre > > > > > draft which is > > > > > worth watching. > > > > > Perhaps if your more specific in what your attempting to do I > > > > > could shed > > > > > more light on the matter. > > > > > > > > > > > > > > > Duane Wright > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > "Ho, Kin Tak" > > > > > > > > > > <ktho@cabletv To: > > > > > <mpls-ops@mplsrc.com> > > > > > .com.hk> cc: > > > > > > > > > > Fax to: > > > > > > > > > > 24/10/2001 Subject: MPLS on > > > > > Foundry and Riverstone > > > > > 05:49 > > > > > > > > > > Please > > > > > > > > > > respond to > > > > > > > > > > ktho > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Does anyone have experience of MPLS for Foundry NetIron > > > and Riverstone > > > > > RS8000? Are they compatible with other vendor ? Any > > > comment on their > > > > > performance ? > > > > > > > > > > ------- > > > > > The MPLS-OPS Mailing List > > > > > Subscribe/Unsubscribe: http://www.mplsrc.com/mplsops.shtml > > > > > Archive: http://www.mplsrc.com/mpls-ops_archive.shtml > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > ************************************************************** > > > > > ************* > > > > > This e-mail is from 186k Limited and is intended only for the > > > > > addressee named above. As this e-mail may contain confidential > > > > > or privileged information, if you are not the named > addressee or > > > > > the person responsible for delivering the message to the named > > > > > addressee, please advise the sender by return e-mail. The > > > > > contents should not be disclosed to any other person > nor copies > > > > > taken. > > > > > 186k Limited is a Lattice Group company, registered in England > > > > > & Wales No. 3751494 Registered Office 130 Jermyn Street, > > > > > London SW1Y 4UR > > > > > http://www.186k.co.uk > > > > > ************************************************************** > > > > > ************* > > > > > > > > > > ------- > > > > > 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 ------- The MPLS-OPS Mailing List Subscribe/Unsubscribe: http://www.mplsrc.com/mplsops.shtml Archive: http://www.mplsrc.com/mpls-ops_archive.shtml
|
|