OSPF Designated Router (DR) and Backup Designated Router (BDR)

Open Shortest Path First (OSPF) uses Designated Routers (DRs) and Backup Designated Routers (BDRs) to optimize performance on multi-access networks like Ethernet. These special roles prevent the network overhead that would occur if every router formed direct relationships with every other router. By centralizing Link-State Advertisement (LSA) exchanges through the DR, OSPF dramatically reduces bandwidth consumption and CPU load while maintaining network reliability.

Key Points

• DRs and BDRs reduce network traffic by limiting full adjacencies to a hub-and-spoke model instead of full mesh
• Elections are based on OSPF interface priority (highest wins) with Router ID as the tie-breaker
• Once elected, a DR is non-preemptive—it won't be replaced by a higher-priority router unless it fails or OSPF resets
• The BDR automatically promotes to DR if the current DR fails, then a new BDR is elected
• All routers maintain an identical Link-State Database (LSDB); the DR only optimizes how updates are distributed
• DR/BDR roles are assigned per broadcast segment, not network-wide

Why OSPF Needs Designated Routers

On broadcast or multi-access networks, forming full adjacencies between every router creates significant problems:

• Bandwidth waste: Each router must exchange LSAs with every other router individually
• CPU overload: Excessive synchronization messages consume processing power
• Slow convergence: Network updates take longer as the number of routers increases
• Scalability issues: The number of adjacencies grows exponentially (O(n²))

The DR solution: By introducing a hub-and-spoke model, non-DR routers form full adjacencies primarily with the DR (and BDR), reducing adjacencies from O(n²) to O(n) while maintaining complete topology awareness.

How DRs and BDRs Work

Core Functions

Centralized Adjacency Formation

• Non-DR routers establish full OSPF adjacencies only with the DR and BDR
• Other routers remain in a 2-WAY state with each other (neighbors but not fully adjacent)
• Dramatically reduces the total number of adjacencies on the segment

Efficient LSA Flooding

• Non-DR routers send LSAs to the DR using multicast address 224.0.0.6
• The DR redistributes LSAs to all routers using multicast address 224.0.0.5
• Prevents redundant transmissions and ensures consistent topology updates

Failure Resilience

• The BDR passively monitors all LSA exchanges and maintains full adjacencies
• If the DR fails, the BDR immediately promotes to DR without disruption
• A new BDR election occurs to fill the vacant backup role

Important Clarifications

• The DR does not make routing decisions—OSPF remains a distributed protocol where each router calculates its own routing table
• All routers maintain their own complete LSDB; the DR only optimizes synchronization
• DR/BDR roles apply per broadcast segment (e.g., per VLAN or Ethernet switch), not across the entire OSPF area

Election Process

Election Criteria

OSPF elects the DR and BDR based on these criteria in order:

1. OSPF Interface Priority (0-255)

   • Default value: 1
   • Higher priority wins
   • Priority 0 makes a router ineligible for DR/BDR roles
   • Configured per interface: ip ospf priority <value>

2. Router ID (RID) (Tie-breaker)

   • Highest Router ID wins if priorities are equal
   • Router ID selection order:
     1. Manually configured RID (router-id command)
     2. Highest IP address on any loopback interface
     3. Highest IP address on any active physical interface

Election Mechanics

Hello Protocol

• Routers exchange Hello packets (default: every 10 seconds on broadcast networks)
• Hello packets contain the router's priority, Router ID, and current DR/BDR information
• New routers learn about existing DR/BDR from received Hello packets

Non-Preemptive Behavior

• A new router with higher priority will not replace an existing DR
• Elections only occur when:
  • No DR currently exists (network initialization)
  • The current DR fails or becomes unreachable
  • The OSPF process is reset or restarted

Example: Router A (priority 100) joins a segment where Router B (priority 1) is already the DR. Router A will not become DR unless Router B fails, even though it has higher priority.

Election Sequence

1. When a segment initializes, all routers wait for the Dead Interval (default: 40 seconds) to discover neighbors
2. The router with the highest priority (and highest RID if tied) becomes DR
3. The router with the second-highest priority becomes BDR
4. All other routers become DROTHERs (neither DR nor BDR)

Failure Scenarios

Scenario	Outcome
DR fails	BDR immediately promotes to DR; new BDR election occurs among DROTHERs
BDR fails	New BDR is elected from DROTHERs; DR remains unchanged
Both DR and BDR fail	Complete new election for both roles
DROTHER fails	No election; only affects that router's adjacencies
OSPF process resets	All adjacencies rebuild; new DR/BDR elections occur

Performance Impact Comparison

Without DR (Full Mesh Adjacencies)

R1 ----- R2
 | \   / |
 |  \ /  |
 |  / \  |
 | /   \ |
R3 ----- R4

Network with 4 routers:

• Each router: 3 full adjacencies
• Total adjacencies: 6
• LSA flooding: Each router sends to 3 others = 12 transmissions per update

With DR/BDR (Hub-and-Spoke)

    R2 (DR)
    /  |  \
  R1   |   R4
    \  |  /
   R3 (BDR)

Same network with DR/BDR:

• DR/BDR: 3 full adjacencies each
• DROTHERs: 2 full adjacencies (DR + BDR)
• Total adjacencies: 5
• LSA flooding: DROTHER sends to DR, DR sends to all = 2 transmissions per update

Enterprise LAN Example

Scenario: 10 routers on a single Ethernet segment

Metric	Without DR	With DR/BDR	Improvement
Adjacencies per router	9	2 (DROTHERs)	78% reduction
Total adjacencies	45	11	76% reduction
LSA transmissions	90 per update	10 per update	89% reduction
Convergence time	High	Low	Faster
CPU load	High	Low	Significant

Common Misconceptions

Misconception	Reality
The DR makes routing decisions for other routers	OSPF is fully distributed; the DR only optimizes LSA flooding. All routers calculate their own routes.
DR/BDR elections are network-wide	Elections occur per broadcast segment. Different segments can have different DRs.
Changing priority triggers immediate re-election	Elections are non-preemptive. Priority changes only take effect after DR failure or OSPF reset.
The DR has a complete view while others don't	All routers maintain identical LSDBs. The DR doesn't have special topology knowledge.
Point-to-point links need a DR	DR/BDR are only used on multi-access networks. Point-to-point links don't elect a DR.

Configuration Best Practices

Priority Assignment Strategy

• **Core/distribution