Redistribution between any pair of routing protocols

Introduction

General information about “Redistribution”:

• Used to take routes from one routing protocol/source and redistribute it into another routing protocol
• Routes must be already in the routing table to be redistributed

Redistribution into EIGRP

General information on “Redistribution from/into EIGRP”:

• Will appear in routing table as “D EX” (EIGRP External)
• Will have an AD of 170

CLI command arguments breakdown:

• source:
  • Mandatory
  • Specifies the source protocol
  • Can be bgp, connected, eigrp, isis, mobile, ospf, static and rip
• metric:
  • Mandatory
  • Specification of bandwidth, delay, reliability, load and MTU
  • A default-metric for all redistributed routes can be defined
  • When a redistributed protocol has a metric specified, the default metric will be overwritten
  • Important: Metric is required since there’s no default metric configured by default!
• route-map:
  • Optional
  • Applies the logic of a route-map to the route while being redistributed

“Redistribution from/into EIGRP” CLI configuration commands:

## Configuring redistribution for EIGRP (classic mode)
Router(config-router)# redistribute <source> [options]

## Defining a default metric for EIGRP (classic mode)
Router(config-router)# default-metric <bandwidth> <delay> <reliability> <load> <mtu>

## Configuring redistribution for EIGRP (named mode)
Router(config-router-af)# topology base
Router(config-router-af-topology)# redistribute <source> [options]

## Defining a default metric for EIGRP (named mode)
Router(config-router-af)# topology base
Router(config-router-af-topology)# default-metric <bandwidth> <delay> <reliability> <load> <mtu>

Redistribution into OSPF

General information on “Redistribution from/into OSPF”:

• External routes will by propagated by Type 5 LSA
• Exception: Inside an NSSA, Type 7 LSA is used!
• The default external metric type is E2
• E1 type routes have metrics that increase as they traverse the network
• E2 type routes have metrics that don’t increase and only use the metric set by the ASBR (20 by default)
• If two E2 routes are learned by the same router it needs to be able to tie break to avoid routing loops
• Tie break when the router and ASBR is in the same area:
  1. Find advertising ASBRs as listed in the Type 5 LSA
  2. Calculate the lowest cost route to reach the ASBR based on the LSDB
  3. Use the outgoing interface and next-hop based on the result
  4. The metric of the route remains unchanged
• Tie break when the router and the ASBR are in different areas:
  1. Calculate the cost to reach the ASBR based on the LSDB
  2. Add the cost from the ABR to the ASBR as listed in a Type 4 LSA
• Important regarding redistribution of OSPF external routes: By default, Cisco IOS only redistributes OSPF Intra-Area and Inter-Area routes. Redistribution of OSPF External (E1/E2) must be explicitly configured!

CLI command arguments breakdown:

• source:
  • Mandatory
  • Specifies the source protocol
  • Can be bgp, connected, eigrp, isis, mobile, ospf, static and rip
• metric:
  • Optional
  • No metric required since there’s a default metric!
  • Routes taken from BGP have a default metric of 1
  • Routes taken from another OSPF process use the source routes metric
  • Routes taken from all other sources use a metric of 20
  • A custom default-metric for all redistributed routes can be defined
  • When a redistributed protocol has a metric specified, the default metric will be overwritten
• metric-type:
  • Optional
  • Can be set to 1 for E1 (External Route 1) type
  • Can be set to 2 for E2 (External Route 2) type
• tag:
  • Optional
  • Used to match the route against a tag
• route-map:
  • Optional
  • Applies the logic of a route-map to the route while being redistributed
• subnets:
  • Optional, but highly recommended
  • Important: Without this command only classful subnets will be redistributed (unique to OSPF)!

“Redistribution from/into OSPF” CLI configuration commands:

## Configuring redistribution for OSPFv2
Router(config-router)# redistribute <source> [options]

## Defining a default metric for OSPFv2
Router(config-router)# default-metric <cost>

## Configuring redistribution for OSPFv3
Router(config-router-af)# redistribute <source> [options]

## Defining a default metric for OSPFv2
Router(config-router-af)# default-metric <cost>

Redistribution into BGP

General information on “Redistribution from/into BGP”

• When redistributing from BGP into an IGP, the command bgp redistribute-internal must be used under the BGP process or either iBGP-learned routes won’t get redistributed (redistribution of iBGP is not recommended)

CLI command arguments breakdown:

• source:
  • Mandatory
  • Specifies the source protocol
  • Can be bgp, connected, eigrp, isis, mobile, ospf, static and rip
• match:
  • Mandatory and only configure for OSPF
  • Specifies which OSPF route types to be redistributed
  • By default, BGP only redistributes OSPF intra-area and inter-area routes
• metric:
  • Optional
  • A custom metric for all redistributed routes can be defined
• route-map:
  • Optional
  • Applies the logic of a route-map to the route while being redistributed

“Redistribution from/into BGP” CLI configuration commands:

## Configuring redistribution for BGP
Router(config-router-af)# redistribute <source> [options]

Preventing Loops with AD

General information on “Preventing Loops with AD”:

• Multiple redistribution points within a routing domain can cause possible routing loops
• One solution is to give redistributed routes a higher AD than non-redistributed routes
• EIGRP does this by default and uses AD 90 for internal routes and AD 170 for external routes
• The distance command can be used to change the AD for each protocol
• The distance command can be linked to an ACL to set per-route AD

“Preventing Loops with AD” CLI configuration commands:

## Modifying the AD for EIGRP (classic mode)
Router(config-router-af)# distance eigrp <eigrp-int-ad> <eigrp-ext-ad>

## Modifying the AD for EIGRP (named mode)
Router(config-router-af)# topology base
Router(config-router-af-topology)# distance eigrp <eigrp-int-ad> <eigrp-ext-ad>

## Modifying the AD for OSPFv2
Router(config-router)# distance ospf intra-area <ospfv2-intra-ad> inter-area <ospfv2-inter-ad> external <ospfv2-external-ad>

## Modifying the AD for OSPFv3
Router(config-router-af)# distance <ospfv3-ad>

## Modifying the AD for BGP
Router(config-router-af)# distance bgp <ebgp-ad> <ibgp-ad> <local-bgp-ad>

Preventing Loops with Route Tags

General information about “Preventing Loops with Route Tags”

• Route tags can be used to prevent routing loops
• Example: Routes from routing domain Domain-A are redistributed into routing domain Domain-B. When entering routing Domain-B the routes are given a tag. If they try to leave the routing domain at another redistribution point back into routing domain Domain-A they can be filtered out based on the tag.