You are on page 1of 45

Which Routing Protocol?

-IPv4 and IPv6 Perspective-

Ciprian Popoviciu

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Public 1
Is one protocol better than the others? Which
routing protocol should I use in my network?
Should I switch from the one I’m using? Do the
same selection rules apply to IPv4 and IPv6?
How will my IPv4 and IPv6 routing protocols
coexist?

IPv4
Ends
Merge
RST-3210 IPv6
11048_05_2005_X2
Presentation_ID © 2005
© 2009 Cisco Systems, Cisco
Inc. Systems,
All rights Inc. All rights
reserved. Ciscoreserved.
Confidential 22
The Questions
 Is one routing protocol better  Converges faster?
than any other protocol?
 Uses less resources?
 Define “Better!”
 Easier to troubleshoot?
 Easier to configure?
 Scales to a larger number of
routers, routes, or neighbors?
 More flexible?
 Degrades more gracefully?
 …

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 3
The Questions
 The answer is yes if:
The network is complex
enough to “bring out” a
protocol’s specific advantages
You can define a specific
feature (or set of features) that
will benefit your network
tremendously…

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 4
The Questions
 But, then again, the
answer is no! 
 Every protocol has
some features and not others,
different scaling
properties, etc.
 Let’s consider some specific
topics for each protocol....

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 5
Before That … The Twist!
 Most likely the IPv6 IGP will  What is the impact on the
not be deployed in a brand convergence of IPv4?
new network and just by itself
 How are the resources shared
 Most likely the existing IPv4 between the two protocols?
services are more important at
 Are the topologies going to be
first since they are generating
congruent?
most of the revenue
 How easy is it to manage
 Redefine “Better!”
parallel IPv4 / IPv6
environments?
 Opportunity to adapt a new
IGP?

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 6
Which Routing Protocol
 IPv4 and IPv6 IGPs
 Convergence Speed
 Design and Topology
Considerations
 Summary

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 7
IPv4 and IPv6 IGPs

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 8
“IPv6 Is an Evolutionary Not a Revolutionary
Step and this is very clear in the case of
routing which saw minor changes even
though most of the Routing Protocols were
completely rebuilt.”

Presentation_ID © 2008
2009 Cisco Systems, Inc. All rights reserved. Cisco Cisco
Confidential
Public 9
The IPv4 – IPv6 Parallel

RIPv2 for IPv4


RIP RIPng for IPv6
Distinct but similar protocols with RIPng taking advantage of IPv6 specificities
OSPFv2 for IPv4
OSPFv3 for IPv6
OSPF Distinct but similar protocols with OSPFv3 being a cleaner implementation
that takes advantage of IPv6 specificities
Extended to support IPv6
IS-IS Natural fit to some of the IPv6 foundational concepts
Supports Single and Multi Topology operation

Extended to support IPv6


EIGRP Some changes reflecting IPv6 characteristics

 For all intents and purposes, the IPv6 IGPs are very similar to their IPv4
counterparts
 IPv6 IGPs have additional features that could lead to new designs

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 10
The Version Agnostic Perspective

 The similarities between the IPv4 and IPv6 IGP lead to


similar network design considerations as far as routing
is concerned
 The implementation of the IPv6 IGPs achieves parity
with the IPv4 counterparts in most aspects but this is an
ongoing development and optimization process
 Coexistence of IPv4 and IPv6 IGPs is a very important
design consideration.

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 11
Convergence Speed

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 12
Convergence Speed
 Equal Cost Convergence
 Link State Convergence
 EIGRP Convergence
 Convergence Summary

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 13
Convergence Speed

 Which protocol converges faster?


 IS-IS vs OSPF vs EIGRP
IS-IS and OSPF have the same characteristics, from a high
level, so we’ll consider them both as link state
Is DUAL faster, or Dijkstra?

 Rules of Thumb
The more routers involved in convergence, the slower
convergence will be
The more routes involved in convergence, the slower
convergence will be

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 14
Convergence Speed

 Five steps to convergence


1. Detect the failure
2. Flood the failure information
3. Calculate new routes around the topology change
4. Add changed routing information to the routing table (RIB)
5. Update the FIB (possibly distributed)

 Steps 1-4-5 are similar for any routing protocol, so we’ll


only look at step 2-3
 But, it’s important to keep them in mind, since they
often impact convergence more than the routing
protocol does

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 15
Equal Cost
 Start with B>C>E and B>D>E
being equal cost A
 If C fails, B and E can shift from
sharing traffic between C and D to B
sending traffic to D only
 Number of routers involved in
convergence: 2 (B and E)
C D
 Convergence time is in the
milliseconds
E

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 16
Link State

 Within a single flooding domain


A single area in OSPF
A single flooding domain in IS-IS

 Convergence time depends on flooding timers, SPF


timers, and number of nodes/leafs in the SPF tree
 What happens when we cross a flooding domain
boundary?

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 17
Link State

 Between flooding domains, link state protocols have


“distance vector” characteristics
 This can have negative or positive impacts on
convergence time in a large network
Reduces tree size
Allows partial SPFs, rather than full SPFs
Introduces translation and processing at the flooding domain
boundaries

 The impact is primarily dependant on the network


design

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 18
Link State – Convergence Data
Tuned IPv4 OSPF, Untuned IPv6 OSPF
 Convergence time with default 2.500

timers and tuned timers 2.000


IPv4 OSPF
1.500

Time
IPv6 OSPF
 IPv4 and IPv6 IGP convergence 1.000 Linear (IPv4

times are similar 0.500


OSPF)
Linear (IPv6
OSPF)
- The IPv6 IGP implementations 0.000
0 500 1000 1500 2000 2500 3000
might not be fully optimized yet Number of Prefixes
- Not all Fast Convergence
optimizations might be available
Tuned IPv4 ISIS, Tuned IPv6 ISIS Tuned IPv4 OSPF, Tuned IPv6 OSPF
0.45 0.5
0.4 0.45
0.35 0.4
IPv4 ISIS 0.35 IPv4 OSPF
0.3
0.3

Time
Time

0.25 IPv6 ISIS IPv6 OSPF


0.25
0.2
Linear (IPv6 0.2 Linear (IPv6
0.15 0.15
ISIS) OSPF)
0.1 Linear (IPv4 0.1 Linear (IPv4
0.05 ISIS) 0.05 OSPF)
0 0
0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000
Number of Prefixes Number of Prefixes

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 19
Link State

 Within a flooding domain


The average convergence time, with default timers, is going to
be in the order of seconds
With fast timers, the convergence time can be in the 100s of
milliseconds
Note: There are operational 200 node IS-IS and OSPF
networks with 500 millisecond convergence times

 Outside the flooding domain


Network design and route aggregation are the primary
determining factors of convergence speed

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 20
EIGRP
 For paths with feasible successors, convergence time is in the
milliseconds
The existence of feasible successors is dependant on the network
design
 For paths without feasible successors, convergence time is
dependant on the number of routers that have to handle and reply
to the query
Queries are blocked one hop beyond aggregation and route filters
Query range is dependant on network design
 Good design is the key to fast convergence in an EIGRP network

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 21
Convergence Summary
IPv4 IGP Convergence Data
 IS-IS with default timers Routes
7000

Milliseconds
 OSPF with default timers
6000
 EIGRP without feasible
successors 5000

 OSPF with tuned timers 4000

 IS-IS with tuned timers 3000

 EIGRP with feasible 2000


successors
Route 1000
Generator

A 0

5000
4000
1000

2000

3000
B C

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 22
Convergence Summary
 It’s possible to converge in under one second using any
protocol, with the right network design
 Rules of Thumb:
More aggregation leads to better performance for EIGRP
Less aggregation leads to better performance for link state
protocols
If you’re going to use link state protocols, tune the timers; but if
you tune the timers, be careful with HA features, like GR/NSF

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 23
The Coexistence Twist
Tuned IPv4 OSPF, Untuned IPv6 OSPF

 IPv6 IGP impact on the IPv4 0.5


0.45
0.4
IGP convergence 0.35 IPv4 OSPF
0.3

Time
 Aggressive timers on both IGPs 0.25
0.2
IPv4 OSPF w/
IPv6 OSPF
0.15
will highlight competition for 0.1 Linear (IPv4
OSPF w/ IPv6
resources 0.05
0 OSPF)
Linear (IPv4
0 500 1000 1500 2000 2500 3000 OSPF)

 Is parity necessary from day 1? Number of Prefixes

Tuned IPv4 ISIS, Tuned IPv6 ISIS Tuned IPv4 OSPF, Tuned IPv6 OSPF
0.6 0.7

0.5 0.6
0.5 IPv4 OSPF
0.4 IPv4 ISIS
0.4

Time
Time

0.3
IPv4 ISIS w/ 0.3 IPv4 OSPF w/
0.2 IPv6 ISIS IPv6 OSPF
0.2
0.1 Linear (IPv4 Linear (IPv4
0.1
ISIS w/ IPv6 OSPF w/ IPv6
0 ISIS) 0 OSPF)
Linear (IPv4 Linear (IPv4
0 500 1000 1500 2000 2500 3000 ISIS) 0 500 1000 1500 2000 2500 3000 OSPF)
Number of Prefixes Number of Prefixes

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 24
Design and Topology
Considerations

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 25
Topology
 Hub and Spoke
 Full Mesh
 Support for Hierarchy
 Topology Summary

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 26
Hub and Spoke Summary

Scaling Issues
No effective means to control
All remote sites receive all other
Link distribution of routing information
remote site link state information;
State moderate scaling capability Care must be taken to prevent
transiting traffic through remote sites
Stub remote routers with filtering and
Care must be taken with summary
EIGRP aggregation; excellent scaling
black holes
capability

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 27
Hub and Spoke
 In the field, we see up to 650
dual homed remotes with 12

EIGRP, and up to about 250 10


with OSPF
8

Seconds
 Tested initial convergence and

EIGRP

EIGRP
OSPF

OSPF
hard failover times 6

600 dual homed remote sites 4

For hard failover, primary hub 2


was powered down
0
Initial Failover
 Testing is still ongoing in this
area

600 remotes

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 28
Full Mesh
Use ip ospf database-filter all out to Manually Designate
OSPF Flooding Points and Increase Scaling Through a Full Mesh
Use isis mesh-group or isis mesh-group blocked to Manually
IS-IS Designate Flooding Points and Increase Scaling Through a
Full Mesh
EIGRP Summarize into and out of the Full Mesh
Summarize

Summarize Summarize

Summarize Summarize

Summarize
New Information
Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 29
Hierarchical Division Points

“Hard” flooding domain, summarization, and filter border; area borders


OSPF need to be considered when designing or modifying the network

“Softer” flooding domain, summarization, and filtering border; L2 overlaps


IS-IS L1 domains, providing some flexibility; network design needs to consider
flooding domain border

Summarization and filtering where configured, no hard or soft borders other


EIGRP than what the network dictates
(But this doesn’t imply the network doesn’t need to be designed!)

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 30
Topology Summary

 Rules of Thumb
EIGRP performs better in large scale hub and spoke
environments
Link state protocols perform better in full mesh environments, if
tuned correctly
EIGRP tends to perform better in strongly hierarchical network
models, link state protocols in flatter networks

Note: With IPv6 a great deal of emphasis is placed on


hierarchical addressing schemes. In certain environments,
EIGRP becomes a good option to support such designs.

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 31
The Coexistence Twist

Multi Process/Topo Single Process/Topo*

• Clear separation of the • Requires less resources


two control planes
• Might provide a more
• Non-congruent deterministic co-
topologies are very existence of IPv4
common if not desired and IPv6
in deployments

*Today most IPv6 IGPs are distinct from their IPv4 counterparts and will run
as ships in the night. The only exception is ISIS.

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 32
Protocol Evolution

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 33
OSPF Future Developments

 Mobile Ad Hoc Networks


 Fast Reroute
Current work in the IETF
draft-bryant-shand-ipfrr-notvia-addresses and others

 Multitopology Routing
draft-ietf-ospf-mt

 Address Families
To support IPv4 and IPv6 in OSPFv3
draft-ietf-ospfv3-af-alt and others

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 34
IS-IS Future Developments

 Fast Reroute
Current work in the IETF
draft-bryant-shand-ipfrr-notvia-addresses and others

 Multitopology Routing
draft-ietf-isis-wg-multi-topology

 Administrative Tags
draft-ietf-isis-admin-tags

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 35
EIGRP Future Developments

 Improved EIGRP hub-and-spoke scaling for DMVP


Significant scaling improvement
IPv4 and IPv6 support

 Based on EIGRP technology: Service Advertisement


Framework (SAF)
Service advertisement / discovery solution based on EIGRP
technology
Does not require EIGRP as an underlying routing protocol (can
be OSPF)
Natively supports both IPv4 and IPv6
First use will be UC Call Agent Discovery

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 36
Summary

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 37
Is one protocol better than the others? Which routing
protocol should I use in my network? Should I switch
from the one I’m using? Do the same selection rules
apply to IPv4 and IPv6? How will my IPv4 and IPv6
routing protocols coexist?

Did we answer this question???

IPv4
Ends
Merge
RST-3210 IPv6
11048_05_2005_X2
Presentation_ID © 2005
© 2009 Cisco Systems, Cisco
Inc. Systems,
All rights Inc. All rights
reserved. Ciscoreserved.
Confidential 38
38
Summary

 There is no “right” answer!


 Consider:
Your business requirements
Your network design
The coexistence between IPv4 and IPv6
Intangibles

 The three advanced IGP’s are generally pretty close in


capabilities, development, and other factors

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 39
Expertise (Intangible)

 What is your team comfortable with?


 What “escalation resources” and other support avenues
are available?
 But remember, this isn’t a popularity contest—you don’t
buy your car based on the number of a given model
sold, do you?
 An alternate way to look at it: what protocol would you
like to learn? 

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 40
Standardization (Intangible)

 Who’s standard?
OSPF: Standardized by the IETF
IS-IS: Standardized by the ISO and the IETF
EIGRP: “Cisco Standard!” 

 Standardization is a tradeoff:
Promises Interoperability
Larger number of eyes looking at problems and finding new
features
Politics often influence standards and causes compromises
New features are often difficult to push through standards
committees, slowing their release

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 41
IPv4/IPv6 Coexistence
 Targeting parity is natural but 450000

consider the tradeoffs during the 400000


350000

Memory (bytes)
early phases of integration 300000
250000
IPv4

IPv6
200000

 IPv4 and IPv6 can be decoupled 150000


100000
Linear
(IPv6)
Linear
offering a unique opportunity to 50000
0
(IPv4)

try a new design with IPv6. Look 0 500 1000 1500 2000 2500 3000

at both congruent and non- Number of Routes

congruent topology approaches


 Evaluate the additional show route afi-all summary

IPv4 Unicast:
resources required by IPv6 ---------------
Route Source Routes Backup Deleted Memory (bytes)
connected 5 1 0 720
local 6 0 0 720
 Take advantage of the IPv6 local SMIAP
static
1
0
0
0
0
0
120
0
ospf 200 3770 1 0 452520
addressing resources! Total 3782 2 0 454080

IPv6 Unicast:
---------------
Route Source Routes Backup Deleted Memory (bytes)
connected 3 1 0 592
local 4 0 0 592
ospf 200 3769 1 0 557960
Total 3776 2 0 559144

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 42
Summary

IP Version Agnostic Rules of Thumb

Mesh Hub and Spoke

Flat Hierarchy
Link
EIGRP
State
Flat Aggregated

Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 43
Recommended Reading
BRKIPM-3301

Source: Cisco Press


Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 44
Presentation_ID © 2009 Cisco Systems, Inc. All rights reserved. Cisco Confidential 45