Transcript Limitations of Layer2 switching

INTRO TO MPLS
CSE 8344 Southern Methodist University
Fall 2003
Introduction



We will be discussing material from MPLS, Technology
and Applications
Read Chapters 1-5 and we will resume this topic after the
MidTerm
These notes will be based on the MPLS book plus some
presentations from Cisco, Nortel, Juniper as well as Dr.
Nair’s notes from last time.
Historical Perspective

X.25, Frame and ATM have proven themselves as a viable Layer2
transport technology
Virtual Circuits are nice for a number of reasons:

–
–
–
–
–

The route through the network is predictable
IP Networks can be built on Top
Just Like Ethernet but spread over the wide area.
The trend has been for the carrier to supply the layer 2 links and the enterprise
to build IP on top. (“Keep the carrier out of my routing.”)
Often the enterprise will get a separate T1 or VC to the internet.
But IP routers are getting much bigger
–
This create lots of problems and questions.
Building big IP Networks

IP is beginning to dominate all other traffic
–

The need is for a cheap scalable IP network is the primary focus
3 Different ways this could be built …..
1) Use physical layer interfaces to connect IP routers
Requires lots of physical layer interface cards in the router$$$$$$$$
Uses lots of bandwidth in backbone, because pipes may be sparsely filled.
2) Use a layer 2 mesh
Use ATM or Frame Switches in the core to create a LOGICAL MESH at Layer 2 with
fewer physical layer connections between routers
OR..
Building big IP Networks (cont)
3)
Build a mesh of routers
This would use routers at both the edge of the network and the core
Use fewer physical layer connections
Lots of routers
Potentially lots of router hops through the network
Routers can be expensive and complicated
Flooded route updates can consume time and bandwidth
DataGram routing only
SO….
Most operators chose to go with Option 2
How can we optimize an IP network
over Layer 2 VCs?

ATM has many features that are NOT necessary to support IP
–
–
–
–

SVC’s not needed
Most of IP has been best effort (so far anyway)
QoS and admission control is overkill
CBR not needed
Early attempts to integrate ATM and IP were interesting:
–
–
LANE, ATMARP, MPOA, etc.
All were attempts to preserve both ATM and IP
Doubly complicated because even more signaling and addressing schemes were
required.
Issues




Price and performance
Scalability
Flexibility of routing functionality
Tight coupling between routing and forwarding algorithms
IP Over ATM
ATM
Issues (IPOATM)

Exponential adjacencies
–
Means that more and more VCs must be built and managed to
support the network.
If not, more hops
–

N^4 messages for topology changes
–

Changing the network can be a headache
Solution
–
Smart Label switching
Extending Router Functionality
D
A
F
C
B
E
Plain old IP routing doesn’t allow us to
route some traffic BD and some BE.
History

Cell switching router (CSR)
• Toshiba
• Mapped the IP signaling to control the ATM Network
• Never really got off the ground
IP switching

–
–
–

Ipsilon
Strip the ATM hardware and use it as a router
Data driven connections through the network
Tag switching
–
–
–
Cisco
External control of Virtual Circuits (LSPs)
Not ATM specific
History (cont)

Multi-protocol Label Switching (MPLS) created in
1997
–
–
–
Consisted of Tag Switching plus other input from
Ipsilon, IBM (ARIS) and others.
Worked in the Internet Engineering Task Force (IETF)
IETF creates the Request For Comments (RFCs) that
are quasi-standards
IP Switching
IP Switching
 Introduced by Ipsilon
 Significant innovations
– General switch management protocol (GSMP)
– Label binding protocol, Ipsilon flow management
protocol (IFMP)
 GSMP allows an ATM switch to be controlled by
an “IP switch controller”
IP Switching Premise
 IP over ATM models are complex and inefficient involve running two control planes
– ATM signaling and routing
– IP routing and address resolution on top
 In contrast IP Switching uses
– IP component plus label binding protocol
– Completely removes ATM control plane
 Goal: To integrate ATM switches and IP routing in
a simple and efficient way
Removing ATM Control Plane
ATM
ARP
IP
MARS
NHRP
PNNI
IP
IFMP
Q.2931
ATM hardware
(a)
(a) IP over Standard ATM
(b) IP Switching
ATM hardware
(b)
IP Switching Architecture
 Switch controller
– Control processor of the system
– Uses GSMP to communicate with ATM switch
– Runs IP routing and forwarding code
 Default VC
– To get control traffic before IP Switching is performed
– Uses well known VCI/VPI value
– Used for data that doesn’t have a label yet
IP Switch Architecture
Switch controller
To
upstream
switch
Default
VC
Data
VC
GSMP
IFMP
Flow Classification and control
Routing
and
forwarding
To
downstream
switch
GSMP
Default
VC
Switch
Data
VC
Switching Basics
 Relies on IP protocols
– To establish routing information
– To determine next hop
 Flow classification and control module selects flows from
incoming traffic
 IP flow refers to a sequence of datagrams
– from one source to one destination, identified by the ordered pair <source
address, destination address>
– can also refer to a flow at finer granularity, e.g., different applications
between same pair of machines, identified by < source address, source port,
destination address, destination port>
Flow Redirection
 Redirection: Process of binding labels to flows and
establishing label switched paths
 Example:
– data is flowing from A via B to C on default VC
– B sends a redirect to A specifying flow y and the label (VPI/VCI) on
which it expects to receive
– If C issues a redirect to B for flow y, B forwards y on the VPI/VCI
specified by C
– Since same flow y enters B on one VC and leaves on another, B uses
GSMP to inform its switching element to set up the appropriate
switching path
Flow Redirection
Redirect:
Flow y VPI/VCI 3/57
A
B
C
Default VC
Default VC
Switch
Controller
Switch
Element
3/57
Switch B issues a REDIRECT message to switch A
Redirect:
Flow y VPI/VCI 3/57
A
Redirect:
Flow y VPI/VCI 2/22
B
Default VC
C
Default VC
Switch
Controller
Switch
Element
3/57
2/22
Switch B and C redirect the same flow, allowing it to be switched at B
Ipsilon Flow Management Protocol
(IFMP)
 Designed to communicate flow to label binding
information
 IFMP is a soft state protocol
 IFMP’s Adjacency Protocol:
– Used to communicate and discover information about neighbors
– Adjacency message sent as limited broadcast
 IFMP’s Redirection Protocol
– Used to send appropriate messages for flow-label bindings
Adjacency Protocol



To exchange initial set of information
ADJACENCY message encapsulated into IP datagram and sent to
limited broadcast address
Also used to agree on the sequence numbers
IFMP’s Redirection Protocol
 Different message types defined:
–
–
–
–
REDIRECT: used to bind label to a flow
RECLAIM: enables label to be unbound for subsequent re-use
RECLAIM ACK: Acknowledgement for RECLAIM message
ERROR: Used to deal with various error conditions
 Common header format
IFMP Redirect Protocol Message Format
Version
Op code
Checksum
Sender Instance
Peer Instance
Sequence Number
Message body: variable length
IFMP REDIRECT message body
Flow type
Flow ID length
Label
Flow identifier
Lifetime
Encapsulation of Redirected Flows
LLC
SNAP IP header
Data
AAL5
trailer
Encapsulation of IP packet on the default VC
IFMP flow
type header
Data
Encapsulation of IP packet on the redirected VCs
AAL5
trailer
General Switch Management
Protocol (GSMP)
 GSMP is a master/slave protocol
– ATM switch is the slave
– Master could be any general purpose computer
 The protocol allows the master to
–
–
–
–
Establish and release VC connections across the switch
Perform port management (Up, Down, Reset, Loopback)
Request Data (configuration information, statistics)
Allows slave to inform master of events such as link failure
GSMP (cont’d)
 GSMP packets are LLC/SNAP encapsulated and sent over
ATM link using AAL5
 GSMP Adjacency Protocol
– Used to gain information about the system at the other end of the
link and
– To monitor link status
 GSMP Connection Management Protocol
– Used to ensure consistency between the GSMP master and slave
– Specifies the QoS using a priority field
Tag Switching
Design Goals
 Adding functionality
– Explicit routing
 Improve scalability
– Hierarchy of routing knowledge
 Link layer independent
– Not just ATM
 Implemented in a variety of devices such as routers and
ATM switches
Terminology Comparison
Terminologies
Tags
Analogies in Label
Switching
Labels
Tag Switching Router (TSR)
Label Switching Router
Tag Edge Router (TER)
Edge Label Switching
Router
Tag Forwarding Information Label Switching Forwarding
Base (TFIB)
Table
Tag Distribution Protocol Label Distribution Protocol
(TDP)
Destination Based Routing


A TSR uses information from unicast routing
protocols to construct its mapping between FECs
and next hops
This mapping is used by the Tag Switching
Control component for constructing the TFIB
which is used for actual packet forwarding
Construction of TFIB
 A local binding between the FEC and a tag
– Takes a tag from the pool of free tags and uses it as an
index in the TFIB to set the incoming tag entry
 A mapping between the FEC and the next hop for
that FEC (provided by the routing protocol(s)
running on the TSR)
 A remote binding between the FEC and a tag that
is received from the next hop
Example
A
B
if0
if1
if2
if1
if2
E
if0
if0
if1
if2
192.6/16
if0
if2
C
if1
D
if0
TSR
Initial TFIB Entries
For FEC
192.6/16
Incoming Outgoing
tag
tag
Next
hop
Outgoing
Interface
On TSR A
100
?
TSR B
If1
On TSR B
6
?
TSR E
If1
On TSR C
17
?
TSR D
If2
On TSR D
5
?
TSR E
If0
On TSR E
6
?
TSR E
If0
TFIB Entries After Tag Distribution
Incoming Outgoing
tag
tag
Next
hop
Outgoing
Interface
On TSR A
100
6
TSR B
If1
On TSR B
6
6
TSR E
If1
On TSR C
17
5
TSR D
If2
On TSR D
5
6
TSR E
If0
On TSR E
6
?
TSR E
If0
Behavior With Routing Change
A
B
if0
if1
if2
if1
if2
E
if0
if0
if1
if2
Link Down
if0
if2
C
if1
D
if0
TSR
Updated TFIB
Incoming Outgoing
tag
tag
Next
hop
Outgoing
Interface
On TSR A
100
6
TSR B
If1
On TSR B
6
6
TSR E
If1
On TSR C
17
5
TSR D
If2
On TSR D
5
6
TSR B
If0
On TSR E
6
?
TSR E
If0
Hierarchical Routing



Scalability
Faster convergence
Fault isolation
Hierarchy of Routing Knowledge
 All TSRs within a routing domain participate in a
common intra-domain routing protocol and
construct TFIB corresponding to destinations within
the domain
 All border TSRs or TERs within a domain and
directly connected TERs from other domains also
exchange Tag binding information via inter-domain
routing protocol
Hierarchy (Cont’d)


To support forwarding,Tag switching allows a
packet to carry several tags organized as a tag
stack
At the ingress, a tag is pushed onto the tag stack,
and at the egress a tag is popped off the stack
Hierarchical Routing Model
Routing
domain
B
V
Routing
domain
C
Routing domain A
T
X
Y
W
Z
TSR
TFIB Entries in Routing Domain A
Incoming Outgoing
tag
tag
Next
hop
On TSR A
N/A
10
TSR X
On TSR B
10
12
TSR Y
On TSR C
12
17
TSR W
On TSR D
17
N/A
TSR W
Label Stack
TSR Z distributes label 2 to TSR W and TSR W gives
label 5 to TSR T for the purpose of inter-domain routing
Top of
Stack
Top of
Stack
10
2
2
Stack after processing in
TSR T
Stack after processing in
TSR W
Multicast in Tag Switching
 Selects the distribution tree based only on
– tag carried in a packet
– interface on which the packet arrives
 TSR maintains its TFIB on a per interface basis
 TSRs connected to a common sub-network agree among
themselves on a common tag associated with a particular
multicast tree
Multicast (Cont’d)
 Partition the set of tags for use with multicast into disjoint
subsets
– Avoid overlap with the help of HELLO packets
 TSR connected to a common sub-network and those
which are a part of the same distribution tree elect one
TSR that will create the tag bindings and distribute them
– any TSR can join the group using the JOIN command
Multicast Model
A
B
TSR
if0
if0
D
if1
if2
if0
E
if0
F
RSVP With Tag Switching
 RSVP supported with the help of a RSVP object the tag Object
 The tag object binding for an RSVP flow carried in
the RSVP “RESV” message
 The RESV message carries the tag object containing
the tag given by a TSR and also information about
the local resources to be used
 The reservation state is refreshed once the flow is set
up using the RESV message
Explicit Routes



Tag switching supports explicit routes with the
help of Explicit Route Object
The object is carried in the RSVP “PATH”
message
The tag information is carried in the Tag Object by
the RSVP “RESV”
Tag Switching Over ATM

VCI field used as tag field
–

For stacks, use VPI
Cell interleave problem
–
–
VC merge
Different tags on the same path
MPLS Terminology
Label Distribution Protocol
LDP, CR-LDP or RSVP
Label Switch Router (LSR)
Label Switch Path
Forwarding
Equivalence
Class
Label Switch Hop
Label Edge Router
(non-standard but useful term)
LER
Forwarding Equivalence
Classes
LSR
LSR
LER
LSP
IP1
IP2
IP1
IP1
#L1
IP1
#L2
IP1
#L3
IP2
#L1
IP2
#L2
IP2
#L3
IP2
Packets are destined for different address prefixes, but can be
mapped to common path
• FEC = “A subset of packets that are all treated the same way by a router”
• In conventional routing, a packet is assigned to a FEC at each hop (i.e. L3
look-up), in MPLS it is only done once at the network ingress
• Mapping a packet to an FEC is known as “classification”.
•This is nominally done via examination of the IP header, but could be
done by other means. (e.g. direct stream adaptation at the LER).
LABEL SWITCHED PATH
(vanilla)
#216
#14
#311
#99
#311
#963
#311
#963
#14
#612
#5
#462
#99
#311
- A Vanilla LSP is actually part of a tree from
every source to that destination (unidirectional).
- Vanilla LDP builds that tree using existing IP
forwarding tables to route the control messages.
IP FORWARDING USED BY HOPBY-HOP CONTROL
Dest
47.1
47.2
47.3
Dest
47.1
47.2
47.3
Out
1
2
3
1 47.1
1
Dest
47.1
47.2
47.3
Out
1
2
3
IP 47.1.1.1
2
IP 47.1.1.1
3
Out
1
2
3
2
IP 47.1.1.1
1
47.2
47.3 3
2
IP 47.1.1.1
• Destination based forwarding tables as built by OSPF, IS-IS, RIP, etc.
Label Switched Path (LSP)
Intf Label Dest Intf Label
In In
Out Out
3
0.50 47.1 1
0.40
Intf Dest Intf Label
In
Out Out
3
47.1 1
0.50
2
2
47.2
2
IP 47.1.1.1
3
1
47.3 3
Label Dest Intf
In
Out
0.40 47.1 1
IP 47.1.1.1
1 47.1
3
1
Intf
In
3
Route=
{A,B,C}
EXPLICITLY ROUTED OR ERLSP
#14
#972
#216
B
#14
A
C
#972
#462
- ER-LSP follows route that source chooses. In
other words, the control message to establish
the LSP (label request) is source routed.
Label Encapsulation
IP | PAYLOAD
“Shim Label” …….
Label VPI VCI
L2
ATM
DLCI
FR
Ethernet
PPP
MPLS Encapsulation is specified over various media types. Top labels may
use existing format, lower label(s) use a new “shim” label format.
MPLS may use the QoS/CoS mechanisms of the media type.