Simulative Comparison of MPLS Protection Switching vs. Ospf Re - Routing

Download as pdf or txt
Download as pdf or txt
You are on page 1of 29

TECHNOLOGY

Simulative Comparison
of
MPLS Protection Switching
CORPORATE

vs.
OSPF Re-routing
Sandrine PASQUALINI
Antoine FROT
Andreas Iselt
Information &
Andreas Kirstädter
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
CONTENTS
TECHNOLOGY

• Introduction
• MPLS
• Label switching
• Recovery
CORPORATE

• OSPF
• Base protocol
• New routing algorithm
• Recovery

• Simulation
• Framework
• Scenarios
• Main results
Information &
Communications
• Conclusion
Networks &
Multimedia
Communications 30 / 09 / 2003
Internet Evolution
TECHNOLOGY

• Issues :
• Increasing amount of multimedia flows over IP
• Need for Quality of Service (QoS)

• Objectives :
CORPORATE

• Future and traditional services on a common IP-platform


• QoS, resilience, scalability

Network Control
Server

Network Admission
Control

IP core
Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
Our objectives
TECHNOLOGY

• Simulate the behaviour of a routing protocol :


• Pre-calculated : MPLS
CORPORATE

• Dynamic and distributed : OSPF


• Resilience : speed up failure detection
improve rerouting

• Compare the resulting recovery speed

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
CONTENTS
TECHNOLOGY

• Introduction
• MPLS
• Label switching
• Recovery
CORPORATE

• OSPF
• Base protocol
• New routing algorithm
• Recovery

• Simulation
• Framework
• Scenarios
• Main results
Information &
Communications
• Conclusion
Networks &
Multimedia
Communications 30 / 09 / 2003
Label Switching
TECHNOLOGY

• Attribution of a label
• On ingress router (Label Edge Router)
CORPORATE

• Not only destination based


• Suppression on egress router
• Packet forwarding
• On transient routers (Label Switch Routers)
• Following a Forwarding Table
• No analysis of packet’s network layer header

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

MPLS Recovery

Global Regional Local


B C B C B C
Protection

A D A D A E F
D
E F E F
CORPORATE

G G G I
H I H I H

P1: Global Protection P2: Regional Protection P3: Local Protection


(by Haskin)
B C B C B C
Restoration

A D A D A D
E F E F E F

G G G
H I H I H I

R1: Global Restoration R2: Regional Restoration R3: Local Restoration


Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
CONTENTS
TECHNOLOGY

• Introduction
• MPLS
• Label switching
• Recovery
CORPORATE

• OSPF
• Base protocol
• New routing algorithm
• Recovery

• Simulation
• Framework
• Scenarios
• Main results
Information &
Communications
• Conclusion
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

OSPF – Main mechanisms

• Hello protocol Thello 10s


detection detection
• Topology changes detection Tdead 4* Thello
• Link state advertisements (LSA) Tlsa 0.6 – 1.1ms
CORPORATE

• Distribution and storage of the topology propagation


Tpropagation
lsaFlood 33ms
• Each router maintains its own view

TSPT O(n.log n), O(n2 )


• Routes calculation
• Distributed, shortest path TSPTdelay 5s

• Forwarding information base (FIB) TSPThold 10s rerouting


rerouting
• Router’s architecture
• Updated when routes have changed TFIB 100 – 300ms

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Routes convergence

• Routes convergence time (RCT)


failure detection, information’s propagation, rerouting 40 – 50s
CORPORATE

• Possible extensions
• Reducing Thello, TSPTdelay et TSPThold

• React faster when resources decrease than when they increase :


• Multipath routing in order to have at least one backup link
• When a link fails, react only locally

expected RCT < 500ms

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

"Outdegree 2" Routing (O2)

• At least 2 outgoing links at each node for each destination


• No single point of failure
• No loops, except "jokers"
CORPORATE

• Minimize: paths’ size/ number of jokers

• Today’s IP-network cores


are sufficiently meshed
• Example : COST 239
"hammoc set" towards Berlin

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Fault tolerance

• Speed up failure detection


• Reduce delays : Thello=100ms, TSPTdelay et TSPThold=0
• Detection at hardware level: 5ms
CORPORATE

• Local reaction (no propagation delays)


• Change of routing tables
• Change of distribution weights

• Preventive mechanism
• After local reaction is completed, recompute routes

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
CONTENTS
TECHNOLOGY

• Introduction
• MPLS
• Label Switching
• Recovery
CORPORATE

• OSPF
• Base protocol
• New routing algorithm
• Recovery

• Simulation
• Framework
• Scenarios
• Main results
Information &
Communications
• Conclusion
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Simulated protocols
• OSPF
• Standard
• Accelerated (Hello or hardware detection)
CORPORATE

• Accelerated, with O2 routing and local reaction

• MPLS
• With a detection mechanism equivalent to OSPF Hellos

B C B C

A
E F D A E F D

G G I
H I H

Information & P2: Regional Protection P3: Local Protection


Communications (by Haskin)
Networks &
Multimedia
Communications 30 / 09 / 2003
Tool Chain
TECHNOLOGY

Demand Topology
(.ndl) (.ndl)

Pathmaker
CORPORATE

MPLS Paths
(.ndl)
New routing
algorithm

Ndl2Tcl

Simulation
(.tcl) NS-2

output
(.nam) NAM

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Simulation Scenarios

• Characteristics
• COST 239 topology
• Fully-meshed traffic
CORPORATE

• One link failure

• Measures
• Sum of all incoming traffic

• Mean of values for each possible link failure

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
User Interface
TECHNOLOGY
CORPORATE

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Comparison of SPF / O2 routing


Overall traffic in the network [Mbytes]
• Just after the failure (2.5s)
• Overall traffic decreases
CORPORATE

• Similar behaviour but


• More traffic is affected in O 2 case
O2
• With equal detection times
• Equal reaction times

Simulation time [s]


Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Restoration time

Processing 0,8
of a LSA
CORPORATE

LSA
flooding
33

Routes 1,20
calculation

Routes 100
installation

0 20 40 60 80 100 (ms)

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Comparison of SPF / O2 routing


Overall traffic in the network [Mbytes]
CORPORATE

The time saved thanks to O2


routing and local reaction is
hidden by TFIB

Simulation time [s]


Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Comparison of MPLS / OSPF

Overall traffic in the network [Mbytes]

• MPLS protection is always faster


• In general (curve a , standard MPLS)
CORPORATE

• With equivalent detection times


(pairs a b , and c d )
100ms less than OSPF

TFIB

Information & Simulation time [s]


Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Conclusion

• Resilience based on traditional OSPF is slow


• Accelerated OSPF :
CORPORATE

• reaches sub-second recovery


• improving restoration speed ?
• speed up failure detection
• but FIB update still an issue

• MPLS protection has superior resilience :


• faster recovery
• but more complex operation, and more bandwidth consuming

Information &
Communications Pre-planned LSP restoration with resource sharing
Networks &
Multimedia
Communications 30 / 09 / 2003
TECHNOLOGY

Thank you for your attention


CORPORATE

Questions ?

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
Global Protection
TECHNOLOGY
Protection
Protection switching,
switching, pre-established
pre-established alternate
alternate LSP,
LSP, global
global scope
scope

C
B
w-LSP
CORPORATE

F
A E D

G G
p-LSP H

+ Single backup LSP per working LSP


– Failure signaling required
+ Node failures covered
Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
Fast Reroute [Haskin]
Protection
Protection switching,
switching, pre-established,
pre-established, pre-reserved,
pre-reserved,
TECHNOLOGY

local
local switching,
switching, global
global recovery
recovery
u Alternative recovery LSP set up from the last-hop LSR in reverse direction to the
ingress LSP and along a node-disjoint path
to the egress LSP

C
CORPORATE

B w-LSP

F
A E D
p-LSP

G G
H

+ Single backup LSP per working LSP + Node failures covered


– No failure signaling required – High spare capacity requirement
Information &
Communications
Networks & Source: [draft-haskin-mpls-fast-reroute-01.txt]
Multimedia
Communications 30 / 09 / 2003
Local Protection
TECHNOLOGY

Protection
Protection switching,
switching, pre-established,
pre-established, local
local scope,
scope, pre-reserved
pre-reserved

C
B
w-LSP
CORPORATE

p-LSP1 F p-LSP3
A E D
p-LSP2

G G
H

– Multiple backup LSPs per working LSP


+ No failure signaling required
– Node failures not covered
Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
Global Restoration
TECHNOLOGY

Restoration,
Restoration, established
established on-demand,
on-demand, reserved
reserved on-demand,
on-demand, global
global scope
scope

C
B
w-LSP
CORPORATE

F
A E D

G G
H

– Failure signaling required + Alternative LSPs distributed over network


+ Node failures covered => high spare capacity efficiency

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
Failure to Egress Restoration
Restoration,
Restoration, pre-established,
pre-established, pre-reserved,
pre-reserved,
TECHNOLOGY

local
local switching,
switching, global
global recovery
recovery

C
B
w-LSP
CORPORATE

F
A E D

G G
H

+ No failure signaling required o Between local and global routing


+ Node failures covered => average spare capacity efficiency

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003
Local Restoration
TECHNOLOGY

Restoration,
Restoration, established
established on-demand,
on-demand, reserved
reserved on-demand,
on-demand, local
local scope
scope

C
B
w-LSP
CORPORATE

F
A E D

G G
H

+ No failure signaling required – Alternative LSPs locally routed


– Node failures difficult to cope with => lower spare capacity efficiency

Information &
Communications
Networks &
Multimedia
Communications 30 / 09 / 2003

You might also like