PPP, ATM, MPLS EECS 489 Computer Networks Z. Morley Mao Monday March 12, 2007 Acknowledgement: Some slides taken from Kurose&Ross

More on Switches r cut-through switching: frame forwarded from input to output port without first collecting entire frame m slight reduction in latency r combinations of shared/dedicated, 10/100/1000 Mbps interfaces

Institutional network hub switch to external network router IP subnet mail server web server

Switches vs. Routers r both store-and-forward devices m routers: network layer devices (examine network layer headers) m switches are link layer devices r routers maintain routing tables, implement routing algorithms r switches maintain switch tables, implement filtering, learning algorithms

Summary comparison

Point to Point Data Link Control r one sender, one receiver, one link: easier than broadcast link: m no Media Access Control m no need for explicit MAC addressing m e.g., dialup link, ISDN line r popular point-to-point DLC protocols: m PPP (point-to-point protocol) m HDLC: High level data link control (Data link used to be considered “high layer” in protocol stack!

PPP Design Requirements [RFC 1557] r packet framing: encapsulation of network-layer datagram in data link frame m carry network layer data of any network layer protocol (not just IP) at same time m ability to demultiplex upwards r bit transparency: must carry any bit pattern in the data field r error detection (no correction) r connection liveness: detect, signal link failure to network layer r network layer address negotiation: endpoint can learn/configure each other’s network address

PPP non-requirements r no error correction/recovery r no flow control r out of order delivery OK r no need to support multipoint links (e.g., polling) Error recovery, flow control, data re-ordering all relegated to higher layers!

PPP Data Frame r Flag: delimiter (framing) r Address: does nothing (only one option) r Control: does nothing; in the future possible multiple control fields r Protocol: upper layer protocol to which frame delivered (eg, PPP-LCP, IP, IPCP, etc)

PPP Data Frame r info: upper layer data being carried r check: cyclic redundancy check for error detection

Byte Stuffing r “data transparency” requirement: data field must be allowed to include flag pattern m Q: is received data or flag? r Sender: adds (“stuffs”) extra byte after each data byte r Receiver: m two bytes in a row: discard first byte, continue data reception m single : flag byte

Byte Stuffing flag byte pattern in data to send flag byte pattern plus stuffed byte in transmitted data

PPP Data Control Protocol Before exchanging network- layer data, data link peers must r configure PPP link (max. frame length, authentication) r learn/configure network layer information m for IP: carry IP Control Protocol (IPCP) msgs (protocol field: 8021) to configure/learn IP address

Virtualization of networks Virtualization of resources: a powerful abstraction in systems engineering: r computing examples: virtual memory, virtual devices m Virtual machines: e.g., java m IBM VM os from 1960’s/70’s r layering of abstractions: don’t sweat the details of the lower layer, only deal with lower layers abstractly

The Internet: virtualizing networks 1974: multiple unconnected nets m ARPAnet m data-over-cable networks m packet satellite network (Aloha) m packet radio network … differing in: m addressing conventions m packet formats m error recovery m routing ARPAnet satellite net "A Protocol for Packet Network Intercommunication", V. Cerf, R. Kahn, IEEE Transactions on Communications, May, 1974, pp

The Internet: virtualizing networks ARPAnet satellite net gateway Internetwork layer (IP): r addressing: internetwork appears as a single, uniform entity, despite underlying local network heterogeneity r network of networks Gateway: r “embed internetwork packets in local packet format or extract them” r route (at internetwork level) to next gateway

Cerf & Kahn’s Internetwork Architecture What is virtualized? r two layers of addressing: internetwork and local network r new layer (IP) makes everything homogeneous at internetwork layer r underlying local network technology m cable m satellite m 56K telephone modem m today: ATM, MPLS … “invisible” at internetwork layer. Looks like a link layer technology to IP!

ATM and MPLS r ATM, MPLS separate networks in their own right m different service models, addressing, routing from Internet r viewed by Internet as logical link connecting IP routers m just like dialup link is really part of separate network (telephone network) r ATM, MPSL: of technical interest in their own right

Asynchronous Transfer Mode: ATM r 1990’s/00 standard for high-speed (155Mbps to 622 Mbps and higher) Broadband Integrated Service Digital Network architecture r Goal: integrated, end-end transport of carry voice, video, data m meeting timing/QoS requirements of voice, video (versus Internet best-effort model) m “next generation” telephony: technical roots in telephone world m packet-switching (fixed length packets, called “cells”) using virtual circuits

ATM architecture r adaptation layer: only at edge of ATM network m data segmentation/reassembly m roughly analogous to Internet transport layer r ATM layer: “network” layer m cell switching, routing r physical layer

ATM: network or link layer? Vision: end-to-end transport: “ATM from desktop to desktop” m ATM is a network technology Reality: used to connect IP backbone routers m “IP over ATM” m ATM as switched link layer, connecting IP routers ATM network IP network

Multiprotocol label switching (MPLS) r initial goal: speed up IP forwarding by using fixed length label (instead of IP address) to do forwarding m borrowing ideas from Virtual Circuit (VC) approach m but IP datagram still keeps IP address! PPP or Ethernet header IP header remainder of link-layer frame MPLS header label Exp S TTL

MPLS capable routers r a.k.a. label-switched router r forwards packets to outgoing interface based only on label value (don’t inspect IP address) m MPLS forwarding table distinct from IP forwarding tables r signaling protocol needed to set up forwarding m RSVP-TE m forwarding possible along paths that IP alone would not allow (e.g., source-specific routing) !! m use MPLS for traffic engineering r must co-exist with IP-only routers

R1 R2 D R3 R4 R A R6 in out out label label dest interface 6 - A 0 in out out label label dest interface 10 6 A D 0 in out out label label dest interface 10 A 0 12 D 0 1 in out out label label dest interface 8 6 A A 1 MPLS forwarding tables

Chapter 5: Summary r principles behind data link layer services: m error detection, correction m sharing a broadcast channel: multiple access m link layer addressing r instantiation and implementation of various link layer technologies m Ethernet m switched LANS m PPP m virtualized networks as a link layer: ATM, MPLS

Outline r Circuit switching m Packet switching vs. circuit switching m Virtual circuits r MPLS m Labels and label-switching m Forwarding Equivalence Classes m Label distribution m MPLS applications r Feedback forms m Fill out during last 20 minutes

Multi-Protocol Label Switching

r Multi-Protocol m Encapsulate a data packet Could be IP, or some other protocol (e.g., IPX) m Put an MPLS header in front of the packet Actually, can even build a stack of labels… r Label Switching m MPLS header includes a label m Label switching between MPLS-capable routers IP packet MPLS header

Pushing, Swapping, and Popping IP Pushing IP Popping IP Swapping r Pushing: add the initial “in” label r Swapping: map “in” label to “out” label r Popping: remove the “out” label R2 R1 R3 R4 MPLS core A B C D IP edge

Forwarding Equivalence Class (FEC) r Rule for grouping packets m Packets that should be treated the same way m Identified just once, at the edge of the network r Example FECs m Destination prefix Longest-prefix match in forwarding table at entry point Useful for conventional destination-based forwarding m Src/dest address, src/dest port, and protocol Five-tuple match at entry point Useful for fine-grain control over the traffic m Sent by a particular customer site Incoming interface at entry point Useful for virtual private networks A label is just a locally-significant identifier for a FEC

Label Distribution Protocol r Distributing labels m Learning the mapping from FEC to label m Told by the downstream router r Example: destination-based forwarding R2 R1 R3 R /24 Pick in-label 10 for /24 I’m using label 10 for /24 I’m using label 43 for /24 In: Link: Out 43: to R4: 10 Map destinations in /24 to out-label 43 and link to R2

Supporting Explicitly-Routed Paths r Explicitly routing from ingress to egress m Set an explicit path (e.g., based on load) m Perhaps reserve resources along the path r Extend a protocol for resource reservation m Start with ReSource Reservation Protocol (RSVP) Used for reserving resources along an IP path m Extensions for label distribution & explicit routing r Extend a protocol for distributing labels m Start with Label Distribution Protocol (LDP) m Extensions for explicit routing & reservation r Two competing proposed standards

Applications of MPLS

TE With Constraint-Based Routing r Path calculation m Constrained shortest-path first m Compute shortest path based on weights But, exclude paths that do not satisfy constraints E.g., do not consider links with insufficient bandwidth r Information dissemination m Extend OSPF/IS-IS to carry the extra information E.g., link-state attributes for available bandwidth r Path signaling m Establish label-switched path on explicit route r Forwarding: MPLS labels

Surviving Failures: Path Protection r Path protection m Reserve bandwidth on an alternate route Protect a label-switched path by having a stand-by m Much better than conventional IP routing Precise control over where the traffic will go Stand-by path can be chosen to be disjoint

Surviving Failures: Fast Reroute r Ensure fast recovery from a link failure m Protect a link by having a protection sub-path r Much faster recovery than switching paths m Affected router can detect the link failure m … and start redirecting to the protection sub-path

BGP-Free Core A B R2 R1 R3 R4 C D /24 eBGP iBGP FEC based on the destination prefix Routers R2 and R3 don’t need to speak BGP

VPNs With Private Addresses A B R2 R1 R3 R4 C D /24 MPLS tags can differentiate pink VPN from orange VPN. Two FECs Direct traffic to orange

Status of MPLS r Deployed in practice m BGP-free core m Virtual Private Networks m Traffic engineering r Challenges m Protocol complexity m Configuration complexity m Difficulty of collecting measurement data r Continuing evolution m Standards m Operational practices and tools

Conclusion r MPLS is an overlay m Tunneling on top of the network Built on top of an underlying routing algorithm m Flexibility in mapping traffic to paths Associating packets with FECs, and then labels m New protocols for creating label-switching tables Binding FECs to labels across a path Establishing explicit routes r Many open questions m Makes operations easier vs. harder? m Trade-offs in exploiting the flexibility? m Interdomain routing with MPLS?