A partner network that only speaks EIGRP
EIGRP (Enhanced Interior Gateway Routing Protocol) and OSPF exchange different routing information. A router running only OSPF cannot read the partner's EIGRP messages as OSPF routes. Redistribution lets R5 take an eligible route learned from another source and advertise it into OSPF.
On small screens, swipe sideways or focus the diagram and use the arrow keys.
The company connects to a partner that runs EIGRP and owns 10.50.0.0/16. R5 speaks both protocols. Users everywhere need to reach the partner, but OSPF routers only know OSPF routes. R5 must redistribute the intended partner route into OSPF, and routers in the normal areas have to work out both the route and how to reach R5.
R5 becomes an autonomous system boundary router (ASBR), the router introducing external reachability. R2 and R4 are area border routers (ABRs), joining area 0 to areas 1 and 2. An area contains detailed OSPF topology records. A link-state advertisement (LSA) is one of those stored records, carried inside an OSPF update packet.
How one partner route travels through three areas
On small screens, swipe sideways or focus the diagram and use the arrow keys.
Device key: LAN 1 (network / site) · R1, R2 (ABR), R3, R4 (ABR), R5 (ASBR) (router) · Partner (EIGRP) (network / service)
Lines show the example’s network connections. Thick links: uplinks/WAN, as labelled. Read the numbered steps for traffic direction and changes.
Reading the addresses
IP labels identify interfaces or destinations. Subnet labels identify networks; /n is the prefix length. A router ID (RID) identifies a routing process and need not be a reachable interface.
An IP address alone is not a URL. A web URL includes a scheme such as https://; an address on a router or printer does not imply a website.
- LAN 1: Subnet 172.16.1.0/24
- R1: RID 1.1.1.1
- R2 (ABR): RID 2.2.2.2
- R3: RID 3.3.3.3
- R4 (ABR): RID 4.4.4.4
- R5 (ASBR): RID 5.5.5.5
- Partner (EIGRP): Subnet 10.50.0.0/16
- 1. R5 learns 10.50.0.0/16 from the partner R5 runs EIGRP towards the partner and redistributes those routes into OSPF. That makes R5 an ASBR.
- 2. R5's type 5 LSA is flooded through normal areas Areas 2, 0 and 1 are normal here. The record says 10.50.0.0/16, metric 20, type E2, advertiser 5.5.5.5, forwarding address 0.0.0.0.
- 3. R4 tells area 0 where R5 is Area 0 has no type 1 LSA for R5, so R4 creates a type 4: ASBR 5.5.5.5 is reachable through me.
- 4. R2 does the same for area 1 R2 creates its own type 4 into area 1, with its cost to reach R5. Now R1 can complete the route.
Why type 5 LSAs cross areas unchanged
A type 5 AS-external LSA describes the imported prefix, mask and external metric. R5 remains its advertising router as it floods through normal areas. ABRs do not originate a replacement type 5 at each boundary, although normal ageing, acknowledgement, refresh and retransmission still happen. Stub and NSSA boundaries do not admit these type 5 records.
On small screens, swipe sideways or focus the diagram and use the arrow keys.
A type 4 ASBR-summary LSA answers a different question: how far is the ASBR from this ABR? R4 describes R5 to area 0, and R2 describes reachability to R5 in area 1. R1 adds its own cost to R2. R1 therefore knows both the external prefix and a usable path towards its originator.
For these metrics, each of the four router-to-router outgoing interfaces from R1 to R5 costs 1. The partner-facing interface on R5 runs EIGRP and is outside OSPF, so the external LSA has forwarding address 0.0.0.0: use R5 as the forwarding target. This field is not a default route; the destination remains 10.50.0.0/16.
What if the forwarding address is not zero?
A non-zero forwarding address names another IP address towards which external traffic should be sent. The calculating router must resolve it using intra-area or inter-area OSPF reachability, not another external route. The internal cost used for E1 or an E2 tie is then towards that address. An external LSA can be present in the database but unusable when this internal reachability is missing.
E2 versus E1 metrics
On small screens, swipe sideways or focus the diagram and use the arrow keys.
With E2, routers still know the internal cost to the ASBR (the forward metric) and use it to compare otherwise eligible E2 paths with the same external metric, for example from two ASBRs. More advanced ASBR-path preference rules can apply before that cost comparison; E2 does not simply mean the nearest exit in every design.
Route type beats metric
For the same prefix, OSPF prefers, in order:
- intra-area (
O), - inter-area (
O IA), - external type 1 (
O E1), - external type 2 (
O E2).
Compare eligible paths and their route type before metrics. An NSSA uses N1/N2 route codes and has additional type 7 preference rules; do not use this normal-area list as a complete NSSA tie-break algorithm.
Redistributing the partner routes on R5
These Cisco IOS/IOS XE style commands begin in global configuration mode. EIGRP process 100 must already have the exact 10.50.0.0/16 route installed on R5. The filter below allows only that prefix; it does not permit every route R5 might learn from the partner. A route map applies the matching rule to redistribution.
On small screens, swipe sideways or focus the diagram and use the arrow keys.
ip prefix-list PARTNER-ONLY seq 5 permit 10.50.0.0/16
route-map PARTNER-TO-OSPF permit 10
match ip address prefix-list PARTNER-ONLY
exit
router ospf 1
redistribute eigrp 100 subnets metric 20 metric-type 2 route-map PARTNER-TO-OSPFR5: import only the installed matching partner prefix, with explicit external seed metric 20 and E2. Prefix-list matching without ge/le is exact; unmatched routes are denied by this route map.
router ospf 1
redistribute eigrp 100 subnets metric 20 metric-type 1 route-map PARTNER-TO-OSPFAlternative on R5 using the same filter: seed 20 and E1. R1 adds internal cost 4, so its total becomes 24. Use this instead of the E2 redistribution line when comparing the cases.
router ospf 1
summary-address 10.50.0.0 255.255.0.0Optional ASBR aggregation example. It combines eligible redistributed component prefixes inside the /16; the exact-/16-only filter above already advertises one prefix, so adding this range does not reduce its route count. area range is used for internal inter-area summaries.
Redistributing in both directions (OSPF into EIGRP and EIGRP into OSPF) on more than one router can create routing loops. Real networks use route maps and tags to control it; that is CCNP material.
Finding the type 4 and type 5 LSAs from R1
On small screens, swipe sideways or focus the diagram and use the arrow keys.
R1#show ip ospf database OSPF Router with ID (1.1.1.1) (Process ID 1) Router Link States (Area 1) Link ID ADV Router Age Seq# Checksum Link count 1.1.1.1 1.1.1.1 802 0x80000006 0x00A3F1 6 2.2.2.2 2.2.2.2 795 0x80000004 0x0051C2 2 Summary Net Link States (Area 1) Link ID ADV Router Age Seq# Checksum 10.0.23.0 2.2.2.2 790 0x80000001 0x00E21B 10.0.34.0 2.2.2.2 790 0x80000001 0x00D93A 10.2.45.0 2.2.2.2 790 0x80000001 0x0077C5 Summary ASB Link States (Area 1) Link ID ADV Router Age Seq# Checksum 5.5.5.5 2.2.2.2 512 0x80000001 0x00F06D Type-5 AS External Link States Link ID ADV Router Age Seq# Checksum Tag 10.50.0.0 5.5.5.5 515 0x80000001 0x00B8D4 0
R1#show ip route ospf | include 10.50 O E2 10.50.0.0/16 [110/20] via 10.1.12.2, 00:08:35, GigabitEthernet0/1
metric-type 1 and the same seed 20 it would read O E1 10.50.0.0/16 [110/24]. The next hop 10.1.12.2 is R2's link address, not R5's router ID.show ip ospf border-routersOn R1: lists R2 as an ABR and 5.5.5.5 as an ASBR reachable through 10.1.12.2.
When external routes go missing or take the long way
- The ASBR has the routes but OSPF doesn't advertise them. No
redistribute, or (on old IOS) nosubnets, so only classful networks went across. - External routes reach some areas but not one. That area is a stub, totally stubby or NSSA area, which blocks type 5 LSAs by design.
- Traffic uses a far exit. Two ASBRs advertise the same prefix as E2 with different seed metrics, so internal distance is ignored. Use E1, or equal seed metrics.
In practice, when an external prefix is present in every area but one, check that area's type before touching the ASBR: a stub flag is the usual answer.
On small screens, swipe sideways or focus the diagram and use the arrow keys.
Three misreadings of E1, E2 and type 4
On small screens, swipe sideways or focus the diagram and use the arrow keys.
- Expecting type 4 LSAs in the ASBR's own area. They are only needed in other areas.
- Thinking E2 routes ignore distance completely; the forward metric still breaks ties.
- Comparing an E1 and an O IA route by metric. The route type decides first.
Rules for external routes
- An ASBR redistributes outside routes as type 5 LSAs, flooded unchanged to every normal area.
- ABRs create type 4 LSAs so routers in other areas can reach the ASBR.
- E2 (default, seed 20 for most redistributed routes, 1 for BGP) keeps the seed metric; E1 adds internal cost.
- For eligible normal-area routes to the same prefix: O, then O IA, then E1, then E2.
Short lab: compare E2 and E1
In the isolated five-router lab, apply the filtered E2 example on R5. On R1, run show ip ospf database external 10.50.0.0 and show ip route 10.50.0.0. Confirm mask /16, advertising router 5.5.5.5, forwarding address zero, external metric 20 and installed E2 metric 20.
Replace only the redistribution line with the filtered E1 alternative. After convergence the route should be E1 metric 24. The seed did not change; the additional 4 is R1's internal cost to R5. Restore the E2 line and confirm metric 20 returns. Test an assigned partner host only after verifying the return route and permitting policy.
Sources: RFC 2328 external path calculation; Cisco forwarding-address explanation. Examples are documentation-based and not validated on a physical router image.
Test yourself on external routes
Which router creates the type 4 LSA that R1 uses to find R5?
R5 redistributes with metric-type 1 and seed 20. R1's cost to R5 is 4. What does R1 show?
A router has O IA 10.50.1.0/24 [110/80] and O E1 10.50.1.0/24 [110/30] for the same prefix. Which is installed?