Build a multi-site addressing plan and advertise one summary per site

Cisco IOSWorking

Lay out a company addressing plan where every branch gets one block that summarizes cleanly, then make each branch router advertise only that one summary into the backbone. R1 is the head office WAN hub in OSPF area 0. Each branch router is an ABR: its WAN link and loopback are in area 0, its LAN subnets are in its own area.

Company plan, carved from 10.0.0.0/8:

Block	Use
10.0.0.0/16	Infrastructure. Loopbacks in `10.0.0.0/24` (/32 each), WAN links in `10.0.1.0/24` (/31 each)
10.1.0.0/16	Head office campus
10.16.0.0/16	Branches 1 to 16, one /20 each
10.17.0.0/16	Branches 17 to 32, reserved

Branch template inside each /20 (shown for branch 1, 10.16.0.0/20; branch 2 is the same layout in 10.16.16.0/20):

VLAN	Use	Branch 1	Branch 2	Gateway
10	Data	10.16.0.0/24	10.16.16.0/24	.1
20	Voice	10.16.1.0/24	10.16.17.0/24	.1
30	Wi-Fi	10.16.2.0/23	10.16.18.0/23	.1
99	Management	10.16.14.0/24	10.16.30.0/24	.1
none	Spare for growth	the rest of the /20	the rest of the /20

Loopbacks: R1 10.0.0.1, BR1 10.0.0.11, BR2 10.0.0.12.

Tasks

Check that each branch's subnets fall inside its /20 and that no two branches overlap.
Address the WAN links as /31s and the loopbacks as /32s from the infrastructure block.
On each branch router, create one subinterface per VLAN with the gateway address.
Run OSPF with the WAN links and loopbacks in area 0 and each branch LAN in its own area. LAN interfaces are passive.
On each ABR, summarize the branch area into one /20.
Confirm R1 sees exactly one route per branch, and that the ABR holds a discard route for its summary.

Solution

1Hub router R1R1

R1 on Cisco IOS

interface Loopback0
 ip address 10.0.0.1 255.255.255.255
!
interface GigabitEthernet0/0/0
 description WAN to BR1
 ip address 10.0.1.0 255.255.255.254
 ip ospf network point-to-point
 no shutdown
!
interface GigabitEthernet0/0/1
 description WAN to BR2
 ip address 10.0.1.2 255.255.255.254
 ip ospf network point-to-point
 no shutdown
!
router ospf 1
 router-id 10.0.0.1
 network 10.0.0.1 0.0.0.0 area 0
 network 10.0.1.0 0.0.0.255 area 0

Some IOS-XE releases print a warning when you put a /31 on an Ethernet interface. It works on a point-to-point link. Setting the OSPF network type to point-to-point skips the DR election on a link that only ever has two routers.

2Branch 1 interfacesBR1

BR1 on Cisco IOS

interface Loopback0
 ip address 10.0.0.11 255.255.255.255
!
interface GigabitEthernet0/0/0
 description WAN to R1
 ip address 10.0.1.1 255.255.255.254
 ip ospf network point-to-point
 no shutdown
!
interface GigabitEthernet0/0/1
 description Trunk to SW1
 no shutdown
interface GigabitEthernet0/0/1.10
 encapsulation dot1Q 10
 ip address 10.16.0.1 255.255.255.0
interface GigabitEthernet0/0/1.20
 encapsulation dot1Q 20
 ip address 10.16.1.1 255.255.255.0
interface GigabitEthernet0/0/1.30
 encapsulation dot1Q 30
 ip address 10.16.2.1 255.255.254.0
interface GigabitEthernet0/0/1.99
 encapsulation dot1Q 99
 ip address 10.16.14.1 255.255.255.0

3Branch 1 OSPF with the area range summaryBR1
BR1 on Cisco IOS
```
router ospf 1
 router-id 10.0.0.11
 passive-interface default
 no passive-interface GigabitEthernet0/0/0
 network 10.0.0.11 0.0.0.0 area 0
 network 10.0.1.0 0.0.0.1 area 0
 network 10.16.0.0 0.0.15.255 area 1
 area 1 range 10.16.0.0 255.255.240.0
```
One network statement with wildcard 0.0.15.255 puts every current and future branch 1 subnet into area 1. area 1 range makes BR1 send one type 3 LSA for 10.16.0.0/20 into area 0 instead of one per subnet, and IOS installs a matching discard route to Null0 on BR1.

4Branch 2 interfacesBR2

BR2 on Cisco IOS

interface Loopback0
 ip address 10.0.0.12 255.255.255.255
!
interface GigabitEthernet0/0/0
 description WAN to R1
 ip address 10.0.1.3 255.255.255.254
 ip ospf network point-to-point
 no shutdown
!
interface GigabitEthernet0/0/1
 description Trunk to SW2
 no shutdown
interface GigabitEthernet0/0/1.10
 encapsulation dot1Q 10
 ip address 10.16.16.1 255.255.255.0
interface GigabitEthernet0/0/1.20
 encapsulation dot1Q 20
 ip address 10.16.17.1 255.255.255.0
interface GigabitEthernet0/0/1.30
 encapsulation dot1Q 30
 ip address 10.16.18.1 255.255.254.0
interface GigabitEthernet0/0/1.99
 encapsulation dot1Q 99
 ip address 10.16.30.1 255.255.255.0

5Branch 2 OSPF with the area range summaryBR2

BR2 on Cisco IOS

router ospf 1
 router-id 10.0.0.12
 passive-interface default
 no passive-interface GigabitEthernet0/0/0
 network 10.0.0.12 0.0.0.0 area 0
 network 10.0.1.2 0.0.0.1 area 0
 network 10.16.16.0 0.0.15.255 area 2
 area 2 range 10.16.16.0 255.255.240.0

Verify

R1 output

R1# show ip route ospf | begin Gateway
Gateway of last resort is not set

      10.0.0.0/8 is variably subnetted, 9 subnets, 3 masks
O        10.0.0.11/32 [110/2] via 10.0.1.1, 00:12:04, GigabitEthernet0/0/0
O        10.0.0.12/32 [110/2] via 10.0.1.3, 00:11:58, GigabitEthernet0/0/1
O IA     10.16.0.0/20 [110/2] via 10.0.1.1, 00:06:40, GigabitEthernet0/0/0
O IA     10.16.16.0/20 [110/2] via 10.0.1.3, 00:06:12, GigabitEthernet0/0/1

One inter-area (O IA) route per branch and none of the /24s or /23s. The 9 subnets are the 4 OSPF routes, the loopback, and the two /31 links with their /32 local routes.

BR1 output

BR1# show ip route ospf | begin Gateway
Gateway of last resort is not set

      10.0.0.0/8 is variably subnetted, 15 subnets, 5 masks
O        10.0.0.1/32 [110/2] via 10.0.1.0, 00:12:04, GigabitEthernet0/0/0
O        10.0.0.12/32 [110/3] via 10.0.1.0, 00:11:58, GigabitEthernet0/0/0
O        10.16.0.0/20 is a summary, 00:06:40, Null0
O IA     10.16.16.0/20 [110/3] via 10.0.1.0, 00:06:12, GigabitEthernet0/0/0

10.16.0.0/20 is a summary ... Null0 is the discard route. A packet for an unused address such as 10.16.9.9 matches it and is dropped here, instead of following a default route back to the hub and looping. Branch 2 also arrives as one /20.

R1 output

R1# show ip route 10.16.2.50
Routing entry for 10.16.0.0/20
  Known via "ospf 1", distance 110, metric 2, type inter area
  Last update from 10.0.1.1 on GigabitEthernet0/0/0, 00:06:40 ago
  Routing Descriptor Blocks:
  * 10.0.1.1, from 10.0.0.11, 00:06:40 ago, via GigabitEthernet0/0/0
      Route metric is 2, traffic share count is 1

A Wi-Fi client address at branch 1 matches the /20 summary, learned from router ID 10.0.0.11 (BR1). Metric 2 is the WAN link (cost 1) plus the summary cost (1).

Mistakes people make

A subnet outside the branch block, for example a printer VLAN numbered 10.16.40.0/24 at branch 1. It falls in branch 3's /20, so it either leaks as its own route or is hidden behind the wrong summary.
Summarizing on the wrong router. area range only works on the ABR that has interfaces in that area. On an internal router it does nothing.
Using summary-address for internal routes. That command is for external (redistributed) routes on an ASBR. Inter-area summaries use area range.
Forgetting the discard route on platforms or protocols that do not add one, which lets traffic for unused addresses loop toward a default route.
Sizing the per-site block for today. A branch that outgrows its /20 needs a second, non-contiguous block and a second route.

On other platforms

The ABR summary on Junos is area-range under the area.

Junos

set protocols ospf area 0.0.0.0 interface ge-0/0/0.0 interface-type p2p
set protocols ospf area 0.0.0.0 interface lo0.0 passive
set protocols ospf area 0.0.0.1 interface ge-0/0/1.10 passive
set protocols ospf area 0.0.0.1 interface ge-0/0/1.20 passive
set protocols ospf area 0.0.0.1 area-range 10.16.0.0/20

The same one-route-per-branch result with EIGRP in classic mode, configured on the WAN interface toward the hub.

Cisco IOS

interface GigabitEthernet0/0/0
 ip summary-address eigrp 100 10.16.0.0 255.255.240.0

Say it in the interview

"I give every site one block that summarizes, here a /20 per branch out of 10.16.0.0/16, and the same VLAN layout inside each block. The branch router is the ABR, so area 1 range 10.16.0.0 255.255.240.0 turns all the branch subnets into one type 3 LSA, and IOS adds a Null0 discard route so unused addresses do not loop. The hub then has one route per branch, and a flapping branch subnet never leaves its area."

Book 3.3 Internet (IP)Book 3.4 Routing