OSPF (Open Shortest Part First)
- What is OSPF(Open Shortest Part First)?
- OSPF (Open Shortest Path First) is:
- A Link-State Routing Protocol
- Interior Gateway Protocol (IGP) – used within an Autonomous System (AS)
- A classless protocol that supports VLSM and CIDR
- Uses cost (based on bandwidth) as its routing metric
- Update are send through multicast IP Address 224.0.0.5 and 224.0.0.6
- OSPF send incremental/Triggered updates.
- Protocol Number: 89
- Uses Dijkstra’s SPF(Shortest Path First) algorithm to calculate the best
path
- OSPF Hello Packets are send every 10 sec and Hold timer is 40 sec.
- Administrative Distance (AD) value is 110.
- Metric = cost = Reference Bandwidth / Interface Bandwidth (Default
Reference Bandwidth = 100 Mbps, You can change it manually to support faster links
like 1Gbps, 10Gbps, etc.) or (10 * 8 / Bandwidth in bps)
- Load balancing via 4 equal cost path by default max up to 16 path can
support (unequal cost load balancing not supported)
- Unlimed Hop count.
- Faster Convergence.
- Hierarchical network design.
- One area has to be designated as Area 0.
- Area 0 is called the backbone area.
- Maintains similar database on all the routers within an area.
- Router ID is used to identify each router.
- Key Features of OSPF:
Feature Description
Link-State Protocol Builds a full map (LSDB) of the network
topology using Link-State Advertisements
Fast Convergence Quickly adapts to changes in the network by
recalculating best paths
Cost-based metric Chooses the best route based on bandwidth
(lower cost = better path)
Support for VLSM/CIDR Fully classless; supports subnetting and
summarization
Authentication support Can use plain-text or MD5 authentication
Multi-area support Supports hierarchical routing for scalability
Uses Dijkstra's SPF algorithm Builds SPF tree to determine shortest path to
every destination
No hop count limitation Unlike RIP, there's no 15-hop limit
- What is Router ID in OSPF?
- Router ID is used to identify the router
- The Router ID (RID) is a 32-bit number (just like an IPv4 address) that
uniquely identifies a router in an OSPF domain (or area).
- Think of it like a name tag for each router in the OSPF network.
- The highest IP assignd to an active physical interface is router ID.
- How is Router ID Selected?
- OSPF automatically chooses the Router ID in this order:
1. Manually configured RID
#router ospf 1
#router-id 1.1.1.1
2. Highest IP address on a loopback interface
- (Loopbacks are preferred because they're always up.)
3. Highest IP address on an active physical interface
- If none are available, OSPF won’t start.
- Important Notes:
- Once RID is chosen, it doesn't change until the OSPF process is restarted
or the router is rebooted.
- After changing the RID manually, you must restart OSPF:
#Router# clear ip ospf process
- How to Check OSPF Router ID:
#Router# show ip ospf
- OSPF Metric Calculation :- In OSPF, the metric is called Cost — it represents
the "expense" or "distance" of sending packets across a particular route.
- Lower cost = better path
- Cost is used to determine the best path to a destination network
- When building the routing table, OSPF adds the costs of all outgoing
interfaces to reach a destination and picks the lowest total cost path.
- How Is OSPF Cost Calculated?
- Cost Formula:
- Cost = Reference Bandwidth / Interface Bandwidth
- Default Reference Bandwidth = 100 Mbps
- So, if you have an interface with:
Bandwidth Cost Formula Resulting Cost
10 Mbps 100 Mbps / 10 Mbps = 10 10
100 Mbps 100 Mbps / 100 Mbps = 1 1
1 Gbps 100 Mbps / 1000 Mbps = 0.1 → Rounded 1
10 Gbps 100 Mbps / 10,000 Mbps = 0.01 → 1 1
- Notice: Anything ≥100 Mbps will have cost 1 by default — not accurate for
modern networks.
- How to Fix Cost Calculation for High-Speed Links?
- Update the Reference Bandwidth:
- Use this command under OSPF router config mode:
#Router(config-router)# auto-cost reference-bandwidth 10000
- This sets the reference bandwidth to 10 Gbps, making the cost values more
meaningful:
Bandwidth New Cost (Ref = 10,000 Mbps)
10 Mbps 10000 / 10 = 1000
100 Mbps 10000 / 100 = 100
1 Gbps 10000 / 1000 = 10
10 Gbps 10000 / 10000 = 1
- How OSPF Works:
1. Neighbor Discovery
- Routers send Hello packets on OSPF-enabled interfaces.
- If both routers agree on parameters (e.g., timers, area ID), they form a
neighbor relationship.
2. Database Exchange
- Routers exchange LSAs (Link-State Advertisements) to learn about each
other’s networks.
- All routers maintain an identical Link-State Database (LSDB) per area.
3. SPF Calculation
- Each router uses Dijkstra's algorithm to build the shortest-path tree.
- From the SPF tree, the router builds the routing table.
- OSPF Packet Types (Total 5):
Packet Type Name Description
1 Hello Discover neighbors and maintain
relationships or adjacency
2 DBD (Database Description) Exchange summaries of LSDBs
3 LSR (Link State Request) Ask for specific LSAs from
neighbors
4 LSU (Update) Send full LSA details to
neighbors
5 LSAck Acknowledge receipt of LSAs
1. Hello Packet
- Used for: Discovering neighbors and maintaining adjacency
- Sent periodically to OSPF-enabled interfaces
- Helps in forming neighbor relationships
- Contains info like:
- Router ID
- Hello/dead intervals
- Area ID
- Router priority
- DR/BDR info
- Important: If two routers’ Hello packets don’t match in key parameters (like
Area ID or Hello timer), they won’t become neighbors.
2. Database Description Packet (DBD or DD)
- Used for: Summarizing the contents of the OSPF LSDB (Link State Database)
- Sent after adjacency is formed
- Contains headers of LSAs (not full info), so routers can compare what LSAs
they are missing
- Think of this as routers saying, “Hey, here’s a list of what I know — tell me
what you’re missing.”
3. Link-State Request Packet (LSR)
- Used for: Requesting specific LSA(s) from a neighbor
- Sent after receiving a DBD packet when a router notices it doesn’t have a
particular LSA
- Example: “Hey, I see you have LSA X, can you send it to me?”
4. Link-State Update Packet (LSU)
- Used for: Sending full LSA information
- Can contain multiple LSAs
- Sent in response to LSR or when topology changes
- Example: “Here’s the full LSA you asked for, or here's an updated route.”
5. Link-State Acknowledgment (LSAck)
- Used for: Acknowledging the receipt of LSU packets
- Ensures reliable delivery of LSAs
- Each LSA must be acknowledged to avoid retransmission
- Like saying: “Thanks! Got it.”
- OSPF Table Types:
Table Name Description
Purpose
Neighbor Table Lists all directly connected OSPF neighbors
For building adjacency
Topology Table (LSDB) Holds the full network topology learned via LSAs
For calculating best paths using SFP algorithm
Routing Table Best paths calculated from SPF tree
For actual forwarding of data packets
1. Neighbor Table
- What it contains:
- List of directly connected OSPF-speaking routers.
- How it's built:
- From Hello packets exchanged on OSPF-enabled interfaces.
- Fields include:
- Neighbor Router ID
- Interface connected
- DR/BDR info
- State (e.g., Init, 2-Way, Full)
- Purpose:
- To manage OSPF neighbor relationships and form adjacencies.
- Without this table, routers can’t communicate to exchange LSAs.
#show ip ospf neighbor
2. Topology Table (LSDB – Link-State Database) or Database Table
- What it contains:
- Complete map of the OSPF area’s topology using LSAs (Link State
Advertisements).
- Built by:
- Exchanging and synchronizing LSAs with neighbors.
- Purpose:
- To represent the entire OSPF network as a graph for SPF (Shortest Path
First) calculation.
- Each router builds an identical LSDB in the area. The Dijkstra algorithm is
run on this database to calculate the best paths.
#show ip ospf database
3. Routing Table
- What it contains:
- Best routes to all known networks based on SPF calculation.
- Populated from:
- The LSDB after running Dijkstra’s algorithm.
- Includes info like:
- Destination network
- Next-hop IP address
- Outgoing interface
- Metric (cost)
- This is the actual table used by the router to forward packets.
- OSPF Area Design (OSPF Hierarchical Structure):
- Area 0 (Backbone Area) is mandatory and must be present.
- All other areas must connect to Area 0.
- Helps reduce the size of LSDBs and limits SPF calculations.
- Improves scalability and performance.
- What is OSPF Area Design?
- OSPF (Open Shortest Path First) is designed to scale efficiently. To do this,
it divides a network into areas.
- An area is a logical grouping of routers. These routers share link-state
information only within that area.
- Why Use OSPF Areas (Hierarchical Design)?
- Here’s why we split OSPF into areas instead of having one big flat network:
Benefit Description
Scalability Smaller LSDB per area → less memory & CPU load
Faster Convergence LSAs are contained in the area, so updates
spread quicker
Reduced Routing Table Size Routes can be summarized between areas
Network Stability Failures in one area don’t affect others as
much
Better Control & Security Each area can be controlled and isolated
logically
- OSPF Hierarchical Structure
- OSPF follows a 2-Level Hierarchical Design:
1. Backbone Area (Area 0):
- The core of the OSPF network
- All other areas must connect to Area 0 (either directly or virtually)
- Think of this as the central highway of OSPF.
2. Non-Backbone Areas (Area 1, 2, 3...):
- These are the normal areas that connect to Area 0.
- Routers in these areas don’t know about every other router in other areas
— they just get summarized routes from Area 0.
- OSPF Router Types in Area Design
- Router Type Description
- Internal Router All interfaces in the same area
- Backbone Router Has at least one interface in
Area 0
- Area Border Router (ABR) Connects one or more non-
backbone areas to Area 0
- Autonomous System Boundary Router (ASBR) Connects OSPF to other routing
protocols (e.g., BGP, RIP)
- Real-World Analogy
- Think of:
- Area 0 = Headquarters (where all big decisions happen)
- Other Areas = Branch offices
- Everyone reports up to HQ, not directly to each other
- What is the Process ID in OSPF?
- The OSPF Process ID is a locally significant number used to identify an OSPF
routing process on a single router.
Router(config)# router ospf 1 (This is the Process ID)
- Locally Significant?
- Yes! It means:
- The Process ID has no effect on other routers.
- You can run multiple OSPF processes on the same router with different process
IDs.
- Routers in the same OSPF network do NOT need to have the same process ID to
form adjacencies.
- Why Do We Use It?
- To differentiate between multiple OSPF processes running on the same router.
- Helps in administrative organization, especially in large-scale or multi-
tenant networks.
- Allows for OSPF process migration or testing without affecting the existing
one.
- What is the Range of Process ID in OSPF?
- The range is: 1 to 65535
- Example:
Router(config)# router ospf 1
Router(config-router)# network 10.0.0.0 0.0.0.255 area 0
Router(config)# router ospf 10
Router(config-router)# network 192.168.1.0 0.0.0.255 area 1
- In this case:
- OSPF process 1 is advertising 10.0.0.0/24 in area 0.
- OSPF process 10 is advertising 192.168.1.0/24 in area 1.
- Both run independently, even though they're on the same router.
- Step-by-Step OSPF Configuration
Step 1: Router(config)# router ospf 1
(1 is the OSPF process ID (can be any number).)
Step 2: Advertise Networks into OSPF
Router(config-router)# network 192.168.1.0 0.0.0.255 area 0
Router(config-router)# network 10.0.0.0 0.0.0.3 area 0
Step 3: Verify OSPF Status
Router# show ip protocols
Router# show ip ospf
Router# show ip ospf neighbor
Router# show ip ospf interface brief
