Back to OSPF Guide Backbone repair

OSPF Virtual Links: Repair a Disconnected Backbone

Learn how a virtual link carries Area 0 through a normal transit area, how to configure both endpoints, and how to prove that the logical backbone path is working.

Area 0Transit AreaABRsRouter IDsBackbone Repair

Virtual Links at a Glance

A virtual link is a logical point-to-point Area 0 connection between two OSPF Area Border Routers. It crosses a normal non-backbone area called the transit area.

Logical connectionArea 0
EndpointsTwo ABRs
Command identifiesRemote router ID
Best roleTemporary repair
Design principle: Use a virtual link to restore service while you repair the topology. A directly connected and redundant Area 0 is easier to understand, monitor, and troubleshoot.

In This Lesson

  1. See the broken-backbone lab
  2. Understand how the virtual link works
  3. Check every requirement
  4. Build the base OSPF topology
  5. Configure both virtual-link endpoints
  6. Verify the logical backbone
  7. Match optional authentication
  8. Troubleshoot a virtual link that stays down
  9. Replace and remove the workaround safely
  10. Complete the guided practice lab
  11. Review the frequently asked questions

1. See the Broken-Backbone Lab

R1 and R2 sit in the main Area 0. R2 reaches R3 through Area 1. R3 connects to another Area 0 segment containing R4. The two Area 0 segments are not physically connected, so the backbone is discontiguous.

Area 0 · Main
R1Backbone router
R2 · ABRRID 2.2.2.2
Area 1 · Transit
Logical Area 0 pathVirtual link
Normal area with full reachability between R2 and R3
Area 0 · Disconnected
R3 · ABRRID 3.3.3.3
R4Backbone router
Configure the virtual link between R2 and R3. Area 1 carries the logical Area 0 adjacency; it does not become part of Area 0.
  • Before repair: R3's Area 0 segment has no continuous path to the main backbone.
  • Temporary repair: R2 and R3 form a logical Area 0 adjacency through Area 1.
  • After repair: R4 can exchange inter-area information through the restored backbone path.

2. Understand How the Virtual Link Works

The virtual link does not create a tunnel interface and does not encapsulate user packets like GRE. It creates an OSPF adjacency that behaves as an unnumbered point-to-point connection in Area 0.

1

R2 finds R3

R2 uses Area 1's intra-area routes to reach router ID 3.3.3.3.

2

Adjacency forms

R2 and R3 exchange OSPF packets as virtual-link neighbors.

3

Area 0 synchronizes

The endpoints exchange backbone LSAs across the logical connection.

4

Routes recover

Inter-area information can cross the repaired backbone again.

Dependency: The virtual link is only as reliable as Area 1. If the transit-area route between the endpoint router IDs disappears, the logical Area 0 adjacency also fails.

3. Check Every Requirement

RequirementWhy it mattersHow to verify
Two endpoint ABRsBoth routers must connect Area 0 to the same transit areashow ip ospf
Matching transit areaBoth commands must name the area between the endpointsshow run | section router ospf
Normal transit areaA stub or NSSA area cannot be used as the transit areaCheck area options on both ABRs
Reachable router IDsEach endpoint needs an intra-area path to the other endpointshow ip route 2.2.2.2
Remote router ID in commandThe command does not use a physical interface addressshow ip ospf
Most common mistake: On R2, enter R3's router ID. On R3, enter R2's router ID. Do not enter the local router ID or a link address.

4. Build the Base OSPF Topology

Set stable router IDs first. Then place the R2–R3 transit link in Area 1 and the outer links in Area 0. The example uses interface-level OSPF commands so the area assignment is easy to read.

R2 · Main backbone endpoint

router ospf 1
 router-id 2.2.2.2
!
interface GigabitEthernet0/0
 ip ospf 1 area 0
interface GigabitEthernet0/1
 ip ospf 1 area 1

R3 · Remote backbone endpoint

router ospf 1
 router-id 3.3.3.3
!
interface GigabitEthernet0/0
 ip ospf 1 area 1
interface GigabitEthernet0/1
 ip ospf 1 area 0
  • Confirm the physical R2–R3 Area 1 adjacency reaches Full.
  • Confirm R2 has an Area 1 path to 3.3.3.3 and R3 has an Area 1 path to 2.2.2.2.
  • Save baseline neighbor, LSDB, and routing-table output before adding the virtual link.

10. Guided Practice Lab

Complete each task in order. Save the important output so you can explain why the link works or fails.

Task 1: Build and Repair the Backbone

  1. Build the four-router topology shown above.
  2. Place the middle R2–R3 link in Area 1.
  3. Confirm the Area 1 adjacency and router-ID reachability.
  4. Configure the virtual link on R2 and R3.
  5. Prove the virtual link and Area 0 adjacency are up.

Expected result: R2 and R3 become Full neighbors through OSPF_VL0, and remote Area 0 routes become available.

Task 2: Create and Diagnose a Safe Fault

  1. Save the working output.
  2. In an isolated lab, enter the wrong remote router ID on R3.
  3. Observe the virtual-link state and neighbor table.
  4. Compare both endpoint configurations.
  5. Restore the correct router ID and confirm recovery.

Expected result: You can identify the endpoint mismatch without clearing the OSPF process.

Task 3: Test the Transit-Area Dependency

  1. Confirm Area 1 carries routes to both endpoint router IDs.
  2. Record the virtual-link cost and physical path.
  3. Shut one lab transit path safely.
  4. Observe whether another intra-area path keeps the link up.
  5. Restore the path and document the dependency.

Expected result: The virtual link survives only when Area 1 still provides an intra-area path between R2 and R3.

11. Frequently Asked Questions

What problem does an OSPF virtual link solve?

It restores logical Area 0 connectivity when a backbone section or non-backbone area cannot connect directly to the main backbone.

Does a virtual link carry normal user traffic inside a tunnel?

No. It creates a logical OSPF adjacency and backbone path. It is not a GRE tunnel and does not create a separately addressed user-data tunnel interface.

Can the transit area be stub or NSSA?

No. The transit area must be a normal area that can provide complete intra-area reachability between the two endpoint router IDs.

Which address belongs in the virtual-link command?

Use the remote ABR's OSPF router ID. Do not use its interface address, the local router ID, or an arbitrary loopback address that is not the active router ID.

Should a virtual link be a permanent design?

Usually no. It is best treated as a temporary repair or migration tool. Replace it with a simple, directly connected, and resilient Area 0 design when possible.

What should I check first when the virtual link is down?

Check that the physical transit-area adjacency is Full and that each endpoint has an intra-area route to the other endpoint's router ID. Then compare the area number, remote router IDs, timers, and authentication.