ospf active backup router

[AllezLom63194]

Hello the community,

Here is my network architecture:

For the moment theses routers are on my desk for network configuration validation.

One headquarter with 2 routers:

RH1 (Main) and RH2 (backup)

each router has 1 LAN, 1 WAN, 1 Back to back interface

These routers are ABR routers connected to area 0 and area 2 (stub)

One site with 2 routers:

RS1 (Main) and RS2 (backup)

each router has 1 LAN, 1 WAN, 1 Back to back interface

These routers are only connected to area 2

The four routers are connected on using WAN interface a one switch on the same VLAN (ID:911 -10.236.1.0/24). So this is a broadcast network.

The back to back connection are OSPF point-to-point network.

The ospf priority are configured like this:

RH1 priority=255

RH2 priority=250

RS1 priority=0

RS1 priority=0

The ospf cost are configured like this:

On all interfaces except bak-to-back that has cost=50

RH1 cost=100

RH2 cost=200

RS1 cost=100

RS1 cost=200

The two headquarter routers RH1 and RH2 send the default route with metric 100 and 200 to the area 2

Here is my problem:

When I issue a traceroute to site LAN network (ex: 10.145.0.0/16) I have this output.

RTR-DC-4501#traceroute 10.145.0.1
Type escape sequence to abort.
Tracing the route to 10.145.0.1
VRF info: (vrf in name/id, vrf out name/id)
1 10.236.1.7 1 msec                      ==============> RS2
   10.236.1.8 0 msec                      ==============> RS1 
   10.236.1.7 0 msec                      ==============> RS2
2 10.145.0.1 1 msec 1 msec 1 msec     =========> LAN connected to RS1 and RS2

My Question:

Do you have any idees why from the headquarter main router I pass thru RS2 then RS1 then RS2 to go to the site LAN ???

Let me know if you need configurations parts.

Thank in advance.

[Elliot Dierksen]

I imagine it has to do with how your OSPF areas are defined. OSPF will always follow intra area routes first, and you can only cross areas via area 0.

[AllezLom63194]

Hi Helliot,

You're maybe wright but I can't find anythink.

The 4 routers are one the area 2 (stub) and only the headquarter routers (RH1 and RH2) are on the area 0.

Router RH1 learn network 10.145.0.0 by the RS2 router with next hop RS1 router but I don't know why.

RTR-DC-4501#sh ip route 10.145.0.0
Routing entry for 10.145.0.0/16
Known via "ospf 1", distance 110, metric 101, type intra area
Last update from 10.236.1.7 (RS2) on GigabitEthernet0/0/3.911, 04:04:35 ago
Routing Descriptor Blocks:
* 10.236.1.8, from 11.11.11.22, 05:05:38 ago, via GigabitEthernet0/0/3.911
Route metric is 101, traffic share count is 1
10.236.1.7, from 11.11.11.21, 04:04:35 ago, via GigabitEthernet0/0/3.911
Route metric is 101, traffic share count is 1

[Elliot Dierksen]

Bear in mind that interfaces are in areas, NOT routers. Try drawing yourself a small diagram of the interfaces and OSPF areas. You can also look at the output of "show ip ospf database" to try and get an understand of the different networks being advertised, but that is usually a lot of information to review.

[AllezLom63194]

Yes I'll do this. 

I have one question to be sur to understand OSPF cost.

RH2 has ip ospf cost = 200 on his WAN interface. So the route to join all networks learned by this router through this WAN interface has the same cost .

So How dose it choose the next hop router RS1 or RS2 ?

[AllezLom63194]

Helliot,

I have drawn the diagram and I can't see, on the RH1 and RH2 router with "sh ip osp database", the router that advertise the network 10.145.0.0.

I can see the route on the RH1 and RH2 router "using sh ip route 10.145.0.0" so somone advertise this network but nothing on the ospf database.

Very strange. Maybe I don't understand the output of this commande...

Regards.

[paul driver]

Hello

To visualise your topology can you share a diagram please highlighting your query

if you are simply querying a traffic flow regards ospf it may be just a metric/cost adjustment that is required on the backup ospf core router 

[AllezLom63194]

Hi,

Here is the diagram of my topology.

I have no ideas for the moment but always look&ing for the origine of the problem.

Thank's for your help.

Regards.

[AllezLom63194]

Hi,

RS1 has a router-id 11.11.11.22

RS2 has a router-id 11.11.11.21

Using this diagram, can you said me witch of RS1 or RS2 router should be the next hop for the network 10.145.0.0 on the RH1 and RH2 router  ?

I don't know what is the selection process when there is equal cost.

Thank in advance,

Regards.

[paul driver]

Hello

I dont see RTR-DC4501 in your topology?

---

@AllezLom63194 wrote:
The two headquarter routers RH1 and RH2 send the default route with metric 100 and 200 to the area 2

---

You have a stub area and are sending a default route so i assume area 0.0.0.2 is a totally stubby area?

How are you sending that a metric with that default route or are you just relying on cost?

By the look of your topology -
RS2 should have a single default  via RS1

RS1 should have three default towards the Dc rtrs

Anything from 10.145.0.0/16 should trraverse first via RS1 then dependig on the destination of the route go via either on of RS1 DC ospf adjecencies

From the DC rtrs taffic should in theroy traverse RS1 first but you say this is not the case

---

@AllezLom63194 wrote:
Do you have any idees why from the headquarter main router I pass thru RS2 then RS1 then RS2 to go to the site LAN ???

---

can you post please the output of those 5 rtrs
sh ip ospf interface brief

[AllezLom63194 · Accepted solution]

Hello Paul,

Thank for your réply.

The RTR-DC4501 is the router with the IP 11.11.11.254

Area 0.0.0.2 is juste à stub area not tatally stubby. I send the metric with the default route.

I have found the solution of my problème.

I have set the max-path to 1 to force only one route on the ASR and set the Router-id with the minus value to the master router on the site.

All works fine.

Thank for the time you have pass on my problem and for your ideas and suggestion

Regards,