In real-world enterprise networks, a single OSPF area cannot scale beyond a few hundred routers due to CPU and memory constraints. Multi-area OSPF solves this by partitioning the network into smaller, manageable areas, reducing LSDB size and speeding up convergence. For the CCNA 200-301 exam (objective 3.4), you must understand the types of routers, LSAs, and the path selection logic in multi-area designs. This chapter gives you the deep, exam-relevant knowledge to design and troubleshoot multi-area OSPF like a pro.
Jump to a section
Imagine a large corporation with a massive headquarters building (the backbone area 0) and several regional branch offices (other areas). Each branch office has its own local network, but all branches must communicate through HQ. The HQ maintains a master directory of all employees across the entire company, but each branch only keeps a directory of its own local employees plus a summary of what's available at other branches. When an employee in Branch A wants to send a package to Branch B, they don't need to know every desk in Branch B—they just need to know the branch's mailing address (the summary route). The HQ's mailroom (Area Border Router, ABR) handles inter-branch routing. If Branch A's internal layout changes (someone moves desks), only Branch A updates its local directory; HQ and other branches only see a summary update. This is exactly how multi-area OSPF works: each area runs its own SPF calculation, ABRs summarize routes between areas, and the backbone area 0 connects everything. The key mechanism is that Type 3 LSAs (summary LSAs) are generated by ABRs to advertise networks from one area into another, while Type 1 and Type 2 LSAs stay within an area. This reduces the LSDB size and CPU load on routers not in the backbone.
What is Multi-Area OSPF and Why Does It Exist?
OSPF (Open Shortest Path First) is a link-state routing protocol that floods link-state advertisements (LSAs) to all routers within an area. In a single area (e.g., area 0), every router has an identical link-state database (LSDB) containing all topology information for the entire network. As the network grows beyond about 50-100 routers, the LSDB becomes huge, consuming memory and CPU for SPF calculations. Convergence time increases because every topology change triggers SPF recalculation on every router.
Multi-area OSPF addresses this by dividing the network into multiple areas, each with its own LSDB. Routers in one area do not know the detailed topology of other areas; they only know summary routes. This dramatically reduces LSDB size and isolates topology changes to a single area. The backbone area (area 0) must connect all other areas—this is non-negotiable. A router that connects an area to the backbone is called an Area Border Router (ABR). Routers that connect to external routing domains (e.g., the Internet) are called Autonomous System Boundary Routers (ASBRs).
OSPF Router Types in Multi-Area
Internal Router (IR): All interfaces belong to the same area. It has a full LSDB for that area only.
Area Border Router (ABR): Has interfaces in at least two areas, one of which must be area 0. It maintains a separate LSDB for each connected area and generates Type 3 LSAs (summary LSAs) to advertise routes between areas.
Autonomous System Boundary Router (ASBR): Injects external routes (e.g., from BGP or static routes) into OSPF. It generates Type 5 LSAs (external LSAs) for routes from outside the OSPF domain.
Backbone Router (BR): A router with at least one interface in area 0. This includes all ABRs and any internal router in area 0.
LSA Types in Multi-Area OSPF
| LSA Type | Name | Generated By | Scope | Description | |----------|------|--------------|-------|-------------| | 1 | Router LSA | Every router | Area | Lists all interfaces and neighbors within the area. | | 2 | Network LSA | Designated Router (DR) | Area | Describes the set of routers attached to a multi-access network. | | 3 | Summary LSA | ABR | Inter-area | Advertises networks from one area into another (prefixes). | | 4 | ASBR Summary LSA | ABR | Inter-area | Advertises the location of an ASBR (used for external route reachability). | | 5 | External LSA | ASBR | Entire AS (except stub areas) | Advertises external routes (Type 1 or Type 2 metrics). | | 7 | NSSA External LSA | ASBR in NSSA | NSSA area only | Used in Not-So-Stubby Areas to carry external routes. |
How Multi-Area OSPF Works Step by Step
Neighbor Discovery: OSPF routers discover neighbors by sending Hello packets on interfaces. In multi-access networks, a Designated Router (DR) and Backup DR (BDR) are elected to reduce LSA flooding.
Database Exchange: After forming a 2-way state, routers exchange Database Description (DBD) packets to compare LSDBs. Missing LSAs are requested via Link State Request (LSR) packets and received as Link State Update (LSU) packets.
LSDB Synchronization: Each router builds its LSDB for the area. ABRs maintain separate LSDBs for each connected area.
SPF Calculation: Each router runs the SPF algorithm on its area's LSDB to compute the shortest path to all networks within the area. For inter-area routes, ABRs inject Type 3 LSAs into the area, which are treated as leaf nodes (no further SPF recursion into the other area).
Route Installation: OSPF installs routes with the lowest metric. Inter-area routes (O IA) have a lower preference than intra-area (O) but higher than external (O E1/E2).
OSPF Area Types
Standard Area: Accepts all LSA types (1,2,3,4,5). Default for area 0.
Stub Area: Blocks Type 5 LSAs (external routes). ABR injects a default route (0.0.0.0/0) via Type 3 LSA. Reduces LSDB size.
Totally Stubby Area: Blocks Type 3, 4, and 5 LSAs. ABR injects a default route. Even smaller LSDB.
Not-So-Stubby Area (NSSA): Accepts Type 7 LSAs (external routes from within the area) but blocks Type 5 LSAs. The ABR translates Type 7 to Type 5 when flooding to other areas.
Totally NSSA: Blocks Type 3, 4, 5 LSAs, but allows Type 7. ABR injects default route.
Key Timers and Defaults
Hello Interval: 10 seconds (broadcast/point-to-point), 30 seconds (NBMA)
Dead Interval: 40 seconds (broadcast/point-to-point), 120 seconds (NBMA)
Wait Timer: Same as Dead Interval (for DR/BDR election)
Retransmit Interval: 5 seconds
Transmit Delay: 1 second
SPF Hold Time: 10 seconds (Cisco default)
Reference Bandwidth: 100 Mbps (Cisco default; can be changed with auto-cost reference-bandwidth)
Cost formula: reference bandwidth / interface bandwidth (in bps)
IOS CLI Verification Commands
show ip ospf
show ip ospf interface
show ip ospf neighbor
show ip route ospf
show ip ospf database
show ip ospf border-routersExample output for show ip ospf neighbor:
Neighbor ID Pri State Dead Time Address Interface
10.1.1.1 1 FULL/DR 00:00:35 192.168.1.2 GigabitEthernet0/0
10.2.2.2 1 FULL/BDR 00:00:37 192.168.1.3 GigabitEthernet0/0Example output for show ip route ospf:
O 10.0.1.0/24 [110/2] via 192.168.1.2, 00:12:34, GigabitEthernet0/0
O IA 172.16.0.0/16 [110/20] via 192.168.1.2, 00:10:01, GigabitEthernet0/0
O E2 0.0.0.0/0 [110/1] via 192.168.1.2, 00:08:22, GigabitEthernet0/0Interaction with Related Protocols
Multi-area OSPF interacts with BGP when redistributing external routes (ASBR). It also works with VRRP/HSRP for first-hop redundancy. In MPLS environments, OSPF can be used as the IGP for label distribution. The key interaction is with the routing table: OSPF routes are compared with other protocols based on administrative distance (OSPF internal = 110, external = 110).
Plan the OSPF Area Hierarchy
Before configuring, decide which routers will be ABRs and which areas will exist. All non-backbone areas must connect to area 0. If a router has interfaces in two non-backbone areas, it cannot be an ABR—you must create a virtual link or redesign. On the CCNA, you might be given a topology and asked to identify ABRs or correct area assignments. Draw the topology: area 0 is the backbone; each other area is a stub or standard area based on requirements.
Configure OSPF on Each Router
Use the `router ospf <process-id>` command to enable OSPF. The process ID is locally significant. Then use `network <network> <wildcard-mask> area <area-id>` to enable OSPF on interfaces. Alternatively, use `ip ospf <process-id> area <area-id>` under the interface. For multi-area, ensure ABRs have interfaces in area 0 and at least one other area. Example: ``` R1(config)# router ospf 1 R1(config-router)# network 192.168.1.0 0.0.0.255 area 0 R1(config-router)# network 10.0.0.0 0.255.255.255 area 1 ``` This makes R1 an ABR.
Verify Neighbor Adjacencies
Use `show ip ospf neighbor` to ensure routers have formed FULL adjacency with their neighbors. For multi-access networks, check for DR/BDR election. If neighbors are stuck in INIT or 2WAY, check Hello/Dead timers mismatch or ACL blocking. Example output: ``` R1# show ip ospf neighbor Neighbor ID Pri State Dead Time Address Interface 10.0.0.2 1 FULL/DR 00:00:32 192.168.1.2 Gig0/0 10.0.0.3 1 FULL/BDR 00:00:34 192.168.1.3 Gig0/0 ``` If you see FULL/ - (dash), it's a point-to-point link.
Check the OSPF Database
Use `show ip ospf database` to view LSAs. In multi-area, ABRs will show Type 3 LSAs in addition to Type 1 and 2. Example: ``` R1# show ip ospf database OSPF Router with ID (10.0.0.1) (Process ID 1) Router Link States (Area 0) Link ID ADV Router Age Seq# Checksum 10.0.0.1 10.0.0.1 1234 0x80000002 0x00A1B2 10.0.0.2 10.0.0.2 1235 0x80000001 0x00C3D4 Summary Net Link States (Area 0) Link ID ADV Router Age Seq# Checksum 10.1.0.0 10.0.0.1 432 0x80000001 0x00E5F6 ``` The Summary Net LSAs are Type 3, injected by the ABR.
Verify Inter-Area Routes in Routing Table
Use `show ip route ospf` to see routes. Inter-area routes are marked with 'O IA'. Example: ``` R1# show ip route ospf O 10.0.1.0/24 [110/2] via 192.168.1.2, 00:12:34, Gig0/0 O IA 172.16.0.0/16 [110/20] via 192.168.1.2, 00:10:01, Gig0/0 ``` The O IA route has a higher metric because it traverses an ABR. If you don't see O IA routes, check that ABR is generating Type 3 LSAs (ensure inter-area summarization is not misconfigured).
Configure Stub or NSSA if Needed
To reduce LSDB size, configure an area as stub or NSSA. On all routers in that area, use `area <area-id> stub` (or `stub no-summary` for totally stubby). For NSSA, use `area <area-id> nssa`. The ABR automatically injects a default route (except for NSSA unless `default-information-originate` is configured). Verify with `show ip ospf` to see area type. Example: ``` R2(config)# router ospf 1 R2(config-router)# area 2 stub ``` Now area 2 will not accept Type 5 LSAs. If you misconfigure only some routers in the area, OSPF adjacency will not form (area type mismatch).
Scenario 1: Large Enterprise Campus with Multiple Buildings
A university campus has a core network (area 0) in the data center, and each building (e.g., Engineering, Science, Library) is a separate area. Each building has its own OSPF area (area 10, 20, 30). The core routers (ABRs) connect each building to area 0. This design isolates failures: if a new switch is added in the Engineering building, only area 10 recalculates SPF. The backbone remains stable. The network engineer configures summarization on the ABRs to reduce the number of Type 3 LSAs into area 0. For example, Engineering uses 10.10.0.0/16, and the ABR advertises only that summary instead of individual /24 subnets. This reduces routing table size and convergence time.
Scenario 2: MPLS VPN with OSPF as the IGP
In a service provider network, OSPF multi-area is used to scale the IGP. The provider edge (PE) routers are in area 0, and customer-facing routers are in separate areas. This allows the provider to isolate customer networks. However, the provider must ensure that the backbone area is contiguous. If a customer site is dual-homed to two different PEs, the PEs must be in area 0 to avoid routing loops. The engineer must carefully design the area boundaries to prevent suboptimal routing.
Common Misconfiguration: Virtual Links
Sometimes a non-backbone area must be connected to area 0 through another non-backbone area, which violates the rule that all areas must connect directly to area 0. The solution is a virtual link, which creates a logical connection through a transit area. However, virtual links are a last resort and add complexity. On the exam, know that virtual links are configured using area <transit-area-id> virtual-link <router-id> on both endpoints. Misconfigured virtual links can cause routing loops or black holes. In production, avoid virtual links by design.
Performance Considerations
In a multi-area OSPF network with thousands of routes, SPF calculation can still be heavy. Engineers often tune SPF timers (timers throttle spf) to increase the delay between SPF calculations. They also use stub areas aggressively to reduce LSDB size. The auto-cost reference-bandwidth command is adjusted to account for high-speed links (e.g., set to 10000 for 10 Gbps). Without adjustment, all Gigabit and 10-Gigabit interfaces have the same cost (1), causing equal-cost multipath (ECMP) and potential suboptimal routing.
What the 200-301 Exam Tests
Objective 3.4: "Configure and verify OSPF multi-area" focuses on:
Identifying router types (IR, ABR, ASBR) based on interfaces and areas.
Understanding LSA types (1,2,3,4,5,7) and their scope.
Configuring and verifying multi-area OSPF (network statements, area assignment).
Interpreting show ip ospf neighbor, show ip route ospf, and show ip ospf database output.
Knowing area types (stub, totally stubby, NSSA) and their effects.
Common Wrong Answers and Why Candidates Choose Them
1. Wrong: "An ABR must have all interfaces in different areas." Right: An ABR must have at least one interface in area 0 and at least one in another area. It can have multiple interfaces in the same non-backbone area. Why: Candidates think "border" means all interfaces are in different areas, but the key is connection to area 0.
2. Wrong: "Type 3 LSAs are generated by ASBRs." Right: Type 3 LSAs are generated by ABRs. ASBRs generate Type 5 LSAs. Why: Confusion between router roles. Remember: ABR = inter-area, ASBR = external.
3. Wrong: "A stub area blocks Type 3 LSAs." Right: A stub area blocks Type 5 LSAs (external), but allows Type 3 (inter-area). Totally stubby blocks Type 3 and 5. Why: Candidates mix up stub and totally stubby. Stub still gets inter-area routes.
4. Wrong: "OSPF multi-area reduces the number of SPF calculations on all routers." Right: It reduces the size of the LSDB and SPF calculation on internal routers, but ABRs must run SPF for each connected area. Why: Oversimplification. ABRs do more work, not less.
Specific Values and Command Outputs
Administrative distance: OSPF internal = 110, external = 110.
Route types: O (intra-area), O IA (inter-area), O E1/E2 (external), O N1/N2 (NSSA external).
Default metric for O E2: 20 (can be changed with default-metric).
In show ip ospf, look for "It is an area border router" or "It is an autonomous system boundary router".
In show ip route ospf, the code 'O IA' is inter-area.
Calculation Traps
Be careful with cost calculation: cost = reference bandwidth / interface bandwidth. If reference bandwidth is changed from default 100 to 1000, a 100 Mbps link costs 10, not 1. On the exam, if a question gives interface speeds and reference bandwidth, calculate the cost. Also, inter-area routes always have a higher cost than intra-area for the same prefix length? Not necessarily—they are just different path types. The metric is additive across areas.
Decision Rule for Scenario Questions
When asked to choose the best area type for a remote site that only needs Internet access: use a totally stubby area. The ABR injects a default route, and the remote router does not learn any inter-area or external routes, minimizing its LSDB. If the remote site has its own external connections (e.g., dual-homed to another AS), use NSSA.
All non-backbone areas must connect directly to area 0; virtual links are a workaround but not recommended.
ABRs generate Type 3 LSAs to advertise routes between areas; Type 5 LSAs are generated by ASBRs for external routes.
Stub areas block Type 5 LSAs; totally stubby areas block Type 3, 4, and 5 LSAs.
OSPF route types: O (intra-area), O IA (inter-area), O E1/E2 (external), O N1/N2 (NSSA external).
Default OSPF cost formula: reference bandwidth (100 Mbps) / interface bandwidth in bps; adjust with auto-cost reference-bandwidth.
Verify multi-area with show ip ospf neighbor (FULL state), show ip route ospf (O IA routes), and show ip ospf database (Type 3 LSAs).
An ABR must have at least one interface in area 0 and one in another area.
These come up on the exam all the time. Here's how to tell them apart.
Single-Area OSPF
All routers have identical LSDB for the entire network.
SPF calculation runs on every router for every topology change.
Scales poorly beyond ~50 routers.
No ABRs or virtual links needed.
All routes are intra-area (O).
Multi-Area OSPF
Each area has its own LSDB; routers only know their area's topology.
SPF calculation is isolated to the area where change occurs.
Scales to hundreds of routers with proper area design.
Requires ABRs to connect areas to area 0.
Routes can be intra-area (O), inter-area (O IA), or external (O E1/E2).
Mistake
An ABR must have interfaces in at least two different non-backbone areas.
Correct
An ABR must have at least one interface in area 0 and at least one in another area. The key is the connection to the backbone.
Candidates think "border" means between any two areas, but the backbone is mandatory.
Mistake
Type 3 LSAs carry the full topology of the source area.
Correct
Type 3 LSAs are summary LSAs that only contain the prefix and metric, not the full topology. This is why inter-area routes are considered 'distance vector' within OSPF.
Candidates assume all LSAs are link-state, but Type 3 is actually a distance-vector advertisement.
Mistake
Stub areas block all external and inter-area routes.
Correct
Stub areas block only Type 5 LSAs (external routes). Inter-area routes (Type 3) are still allowed. Totally stubby blocks both Type 3 and Type 5.
The term 'stub' is misinterpreted as 'isolated', but it only restricts external routes.
Mistake
The OSPF process ID must match on all routers in the same area.
Correct
The process ID is locally significant and does not need to match between routers. OSPF neighbor formation uses the Router ID, not the process ID.
Candidates confuse process ID with area ID or autonomous system number.
Reveal each answer, then mark whether you got it right. Score 60%+ to unlock the next chapter.
No, by definition an ABR must have at least one interface in area 0. The term 'Area Border Router' implies it borders the backbone area. If a router has interfaces in two non-backbone areas, it is not an ABR and those areas cannot communicate through it without a virtual link through area 0. On the exam, if you see a router with interfaces in area 1 and area 2, it is not an ABR; it is just an internal router for both areas, but OSPF will not route between them unless area 0 is involved.
O E1 (External Type 1) routes include the internal cost to the ASBR plus the external metric. O E2 (External Type 2) routes use only the external metric; the internal cost to the ASBR is not added. By default, OSPF redistributes routes as E2 with a metric of 20. E1 is preferred over E2 when comparing routes from the same ASBR because it reflects the total path cost. On the exam, if you see two external routes to the same destination, the one with lower total cost (E1) is preferred.
To configure a totally stubby area, use the command `area <area-id> stub no-summary` on the ABR. On all other routers in that area, use `area <area-id> stub` (without no-summary). The ABR will then block Type 3, 4, and 5 LSAs and inject a default route (0.0.0.0/0) into the area. This drastically reduces the LSDB size. Verify with `show ip ospf` to see the area type as 'Stub no summary'.
A virtual link is a logical connection between two ABRs through a non-backbone area (transit area) to extend the backbone area. It is used when a non-backbone area cannot be physically connected to area 0, such as when merging two OSPF domains. Virtual links are configured using `area <transit-area-id> virtual-link <router-id>` on both endpoints. They are a last resort because they add complexity and can cause routing issues. On the CCNA exam, know that virtual links exist and their configuration syntax, but they are rarely tested in depth.
On multi-access networks (e.g., Ethernet), a router may remain in 2WAY state with a neighbor that is not the DR or BDR. This is normal—routers form FULL adjacency only with the DR and BDR. They maintain 2WAY with all other routers to receive LSAs via the DR. If you see 2WAY on a point-to-point link, it indicates a problem (e.g., mismatched Hello timers). Check interface type with `show ip ospf interface`.
The OSPF router ID is the highest loopback IP address, or if no loopback, the highest active physical interface IP. To set it manually, use `router-id <ip-address>` under the OSPF process. After changing, you must reload OSPF with `clear ip ospf process` or reload the router. On the exam, know that the router ID must be unique across the OSPF domain.
Both stub and NSSA block Type 5 LSAs, but NSSA allows Type 7 LSAs to carry external routes that originate within the NSSA. The ABR translates Type 7 to Type 5 when flooding to other areas. Stub areas do not allow any external routes (Type 5 or Type 7). Use NSSA when a remote site has its own external connection (e.g., a backup Internet link) but you still want to reduce LSDB size. On the exam, if a question says a remote site needs to advertise a default route to the rest of the OSPF domain, NSSA is the correct choice.
You've just covered OSPF Multi-Area Design — now see how well it sticks with free CCNA 200-301 practice questions. Full explanations included, no account needed.
Done with this chapter?