Are received RIB-failed BGP routes redistributed into an IGP?

[Sam Brynes]

We have a lab set up where two BGP routers R1 and R2 are peered up with each other using iBGP and these two BGP routers are peered up with BGP routers in another AS (eBGP). We are also running an IGP (OSPF) on these two routers (R1 and R2), as well as on all routers within our test organization TestLab.

On R1 we are using a route-map to set the local preference of the received route to subnet 1.0.0.0/24 to 500, and on the other BGP router R2 we are using a route-map to set the local preference of the received route to subnet 2.0.0.0/24 to 500.

We have configured our routes received by eBGP (1.0.0.0/24 and 2.0.0.0/24) to be redistributed into our IGP (OSPF) on both R1 and R2. On R1, the subnet 2.0.0.0/24 shows up with a RIB failure because it is receiving the same subnet via our IGP (OSPF).

My question is if the 2.0.0.0/24 subnet on R1, which is received from R2 via iBGP is still redistributed into the IGP (OSPF)?

[Richard Burts]

Sam

You have given us a somewhat incomplete description of what is going on in this situation and that makes it difficult to give you a really good answer.

You have told us that R1 gets a RIB failure for network 2.0.0.0/24. You have not told us whether R2 has the same RIB failure. And you have not told us how 2.0.0.0/24 gets into OSPF. Knowing  these things would be important in finding the right answer for your question.

HTH

Rick

[milan.kulik]

Hi Richard,

I agree the problem description is not clear.

But anyway:

AFAIK, a prefix must be present in RIB to be allowed to redistribute to other routing protocol, correct?

So if "On R1, the subnet 2.0.0.0/24 shows up with a RIB failure" it can't be redistributed to OSPF.

It could only be advertised to BGP neigbors in a case the "bgp suppress-inactive" command was not used, see

http://www.cisco.com/en/US/docs/ios/iproute_bgp/command/reference/irg_bgp1.html#wp1116464

for details.

Best regards,

Milan

[Richard Burts]

Milan

Clearly the RIB failure on R1 is because the route was present in OSPF. And I agree that R1 will not redistribute a route learned from BGP unless BGP has successfully inserted that route into the RIB. So R1 will not do redistribution of the route from BGP into OSPF. My questions were trying to find out if the route is in OSPF because it was redistributed by R2 or if it was in OSPF from some other source.

HTH

Rick

[Sam Brynes]

Rick,

Both subnets 1.0.0.0/24 and 2.0.0.0/24 are configured to be redistributed from BGP into OSPF on both BGP routers R1 and R2. No other router is doing any redistribution from BGP into OSPF.

Sam

[Giuseppe Larosa]

Hello Sam,

taken on R1 and R2 post the following

show ip ospf database external 1.0.0.0

show ip ospf database extenal 2.0.0.0

look for the advertising router

As explained by Rick what you see may be normal behaviour in a redundant scenario.

Hope to help

Giuseppe

[Richard Burts]

Sam

I have been trying to figure out whether R2 is successfully learning the route to 2.0.0.0/24 and successfully redistributing it into OSPF. So far I have not got a clear answer about this.

If it is true that R2 is learning 2.0.0.0/24 via BGP and is successfully redistributing it into OSPF then the RIB failure on R1 is an expected behavior. In that case everything is working as expected. And to answer your specific question - no R1 will not redistribute 2.0.0.0/24 into OSPF if there is RIB failure on R1 for 2.0.0.0/24.

HTH

Rick

[Sam Brynes]

Rick,

How would we know if "R2 is successfully learning the route to 2.0.0.0/24 and successfully redistributing it into OSPF"? What output would we need to use to figure out this information?

Sam

[Richard Burts]

Sam

From the output of show ip ospf database external from R1 for the 2.0.0.0 entry we get this

Advertising Router: 10.0.2.1

So pretty clearly R1 is learning the route from R2

And from the output on R2 for 2.0.0.0 we get this

Advertising Router: 10.0.2.1

which suggests that R2 redistributed the route.

There are a couple more outputs which should help to prove the point.

- can you post the output of show ip bgp 2.0.0.0 from both R1 and R2?

- can you post the output of show ip route 2.0.0.0 from both R2 and R1?

We are looking in particular to see if the route on R2 is a BGP route and if the next hop is the external BGP neighbor.

HTH

Rick

[Sam Brynes]

Here are the outputs from R1 and R2:

R1#show ip ospf database external 1.0.0.0

            OSPF Router with ID (10.0.5.2) (Process ID 7)

                Type-5 AS External Link States

  LS age: 1401

  Options: (No TOS-capability, DC)

  LS Type: AS External Link

  Link State ID: 1.0.0.0 (External Network Number )

  Advertising Router: 10.0.5.2

  LS Seq Number: 80000001

  Checksum: 0xA4FB

  Length: 36

  Network Mask: /24

        Metric Type: 2 (Larger than any link state path)

        TOS: 0

        Metric: 1

        Forward Address: 0.0.0.0

        External Route Tag: 1

R1#show ip ospf database external 2.0.0.0

            OSPF Router with ID (10.0.5.2) (Process ID 7)

                Type-5 AS External Link States

  Routing Bit Set on this LSA

  LS age: 1390

  Options: (No TOS-capability, DC)

  LS Type: AS External Link

  Link State ID: 2.0.0.0 (External Network Number )

  Advertising Router: 10.0.2.1

  LS Seq Number: 80000003

  Checksum: 0xAEF2

  Length: 36

  Network Mask: /24

        Metric Type: 2 (Larger than any link state path)

        TOS: 0

        Metric: 1

        Forward Address: 0.0.0.0

        External Route Tag: 1

R2#show ip ospf database external 1.0.0.0

            OSPF Router with ID (10.0.2.1) (Process ID 7)

Type-5 AS External Link States

  Routing Bit Set on this LSA

  LS age: 1622

  Options: (No TOS-capability, DC)

  LS Type: AS External Link

  Link State ID: 1.0.0.0 (External Network Number )

  Advertising Router: 10.0.5.2

  LS Seq Number: 80000001

  Checksum: 0xA4FB

  Length: 36

  Network Mask: /24

Metric Type: 2 (Larger than any link state path)

TOS: 0

Metric: 1

Forward Address: 0.0.0.0

External Route Tag: 1

R2#

R2#show ip ospf database external 2.0.0.0

            OSPF Router with ID (10.0.2.1) (Process ID 7)

Type-5 AS External Link States

  LS age: 1614

  Options: (No TOS-capability, DC)

  LS Type: AS External Link

  Link State ID: 2.0.0.0 (External Network Number )

  Advertising Router: 10.0.2.1

  LS Seq Number: 80000003

  Checksum: 0xAEF2

  Length: 36

  Network Mask: /24

Metric Type: 2 (Larger than any link state path)

TOS: 0

Metric: 1

Forward Address: 0.0.0.0

External Route Tag: 1

R2#