External default route sticking in routing table

[#1]

Hi, I'm currently studying computer networks and have run into a bit of an issue.

 

Short description of problem: A backup default-route on R4 made available through IP SLA+tracking of an interface going down (on R3), then shared with default-information originate into OSPF won't remove itself from the OSPF-network when tracking deactivates the default-route on R4 (from where it is shared). I've tried setting the metric when originating to super-high values. Why does this happen?

In a recent lab we redistributed an EIGRP-network into OSPF (only EIGRP into OSPF, we used a summary static route for traffic from EIGRP ---> OSPF). We only redistributed EIGRP into OSPF from R3. R3 also redistributed the aforementioned static summary to the EIGRP-network. Therefore, traffic between the networks would go over R3. In case any of R3s links were to fail we had to configure backup static routes on R4. I chose to do this by using SLA and tracking on both of R3s interfaces tracking when any of them went unavailable. It looked like this (a notepad with the configs is attached btw):

R4

.....

ip sla 1
icmp-echo 15.0.0.70
timeout 1000
threshold 2
frequency 5
exit
!
ip sla schedule 1 life forever start-time now
!
ip sla 2
icmp-echo 15.0.0.81
timeout 1000
threshold 2
frequency 5
exit
!
ip sla schedule 2 life forever start-time now
!
track 1 rtr 1 reachability
track 2 rtr 2 reachability
!
track 3 list boolean or
object 1 not
object 2 not
!
ip route 15.0.0.80 255.255.255.240 s0/0 track 3
ip route 0.0.0.0 0.0.0.0 s0/1 track 3
!
ip access-list standard TO_EIGRP
permit 15.0.0.80 0.0.0.15
exit
!
route-map EIGRP
match ip address TO_EIGRP
!
!
router eigrp 1
no auto-summary
passive-interface default
redistribute static route-map EIGRP
no passive-interface s0/1
network 15.0.0.73 0.0.0.0
exit
!
router ospf 1
passive-interface default
default-information originate
no passive-interface s0/0
network 15.0.0.85 0.0.0.0 area 0
 

When any of the links on R3 are shut down the static routes on R4 are redistributed to where I want (OSPF-network summary to the EIGRP-network and default-route to OSPF), but when the the links on R3 are brought up again the default-route redistributed into the OSPF-network does not remove itself from the routing table (ie on R5) despite the route disappearing from R4s table. Why is this happening? And what can I do to fix it?

For the purpose of the lab this is a non-issue as giving the default-route a high enough metric that the path through R3 is taken instead was enough, but I'd like to understand why this is happening so I can avoid it in the future/fix it if it ever happens again. I hope this hasn't been too hard to read and understand.

Thanks in advance for any help,

Pat

[Jon Marshall]

Pat

You didn't include all the configs but I got the general idea.

I tested it on an emulator (not GNS3) and it worked fine for me.

Can you just confirm something for me.

When you shut down the interface on R3 to R5 does R5 immediately know it's end is down ?

It's been a while since I used GNS3 and I can't remember. If it does realise it has gone down then it should also realise the OSPF neighborship is down.

The emulator I am using doesn't realise this and there is a big difference in the failover times depending on which interface you shut down on R3 which I can explain but there isn't much point if R5 in your scenario realises the other end of the link has gone down.

Basically it all depends on how quickly R5 realises it's has lost it's OSPF neighborship with R3.

Jon