↓ Skip to main content
KAZULOG.INFO
  1. Network Articles/
  2. OSPF Articles/

OSPF Network Types

Table of Contents

What is an OSPF network type

OSPF changes its behaviour according to the nature of the link an interface is attached to. That classification is the network type. A link like Ethernet, where any number of routers can share one segment and multicast is delivered as-is, a leased line that connects exactly two routers, and a Frame Relay cloud where multicast does not get through all need different ways of finding neighbors and different answers to the question of whether a DR/BDR should be elected.

The network type changes the following four things.

What changesDescription
Hello / dead intervalHow often Hellos are sent, and how long after Hellos stop arriving a neighbor is declared down
DR/BDR electionOn types where three or more routers can share a segment, a DR/BDR is elected to cut down the number of adjacencies (DR and BDR)
How neighbors are foundAutomatically via multicast (224.0.0.5), or statically with the neighbor command
Representation in the LSDBTypes that elect a DR/BDR produce a Network-LSA (Type 2) and the link appears as Transit in the Router-LSA. Types that do not produce no Network-LSA, and the link is advertised as Point-to-Point

The Hello and dead intervals have to match between neighbors. A Hello whose intervals do not match is discarded, so two ends with different network types can fail to form an adjacency for that reason alone.

The list of network types

RFC 2328 defines four network types (Section 1.2). Cisco’s implementation adds one more.

TypeHello / DeadDR/BDRFinding neighborsNetwork-LSATypical use
Broadcast10 / 40 sElectedAutomatic (224.0.0.5)YesEthernet (the default)
Point-to-point10 / 40 sNot electedAutomatic (224.0.0.5)NoA link with exactly two routers
Non-broadcast (NBMA)30 / 120 sElectedStatic (neighbor)YesA full mesh over Frame Relay or ATM
Point-to-multipoint30 / 120 sNot electedAutomatic (224.0.0.5)NoA partial mesh
Point-to-multipoint non-broadcast30 / 120 sNot electedStatic (neighbor)NoA partial mesh without multicast (a Cisco extension, not one of the four in the RFC)

show ospf interface also shows LOOPBACK for loopback interfaces. That is not a link on which Hellos are sent; it only appears in the Router-LSA as a host route (a /32 stub network).

Verification on real devices

Three Cisco IOS XR (XRd 26.1.1) routers are connected in a line with every link in area 0, and only the network type of the R1–R2 Ethernet link is changed. The R2–R3 link is left at the default broadcast, so two types can be observed at the same time on R2.

On IOS XR the network type is set with the network command on the interface under router ospf. The default on Ethernet is broadcast.

The test runs in eight stages (STEP 0 to 7). The show output, syslog and running-config collected from all three routers at every STEP are gathered in Verification config and show output at the end of the article.

STEPOperationResult on the R1–R2 link
0Initial state (the default)BROADCAST, Hello 10/Dead 40, a DR/BDR is elected
1network point-to-point on both endsPOINT_TO_POINT, no DR/BDR, the Network-LSA disappears
2Put R1 back to the default (mismatched types)No adjacency forms and every OSPF route is lost
3Put both ends back to the defaultRecovered
4network non-broadcast on both endsNON_BROADCAST, Hello 30/Dead 120, no neighbor is discovered
5Configure neighbor staticallyThe adjacency forms and Hellos become unicast
6network point-to-multipoint on both endsPOINT_TO_MULTIPOINT, Hello 30/Dead 120, no DR, no neighbor needed
7Remove the network configuration (final state)Restored to the initial state

Broadcast is the default (STEP 0)

show ospf interface shows the network type, the DR/BDR and the timers in one place. This article uses it as the main tool.

STEP 0 R1 the default (broadcast)
GigabitEthernet0/0/0/0 is up, line protocol is up 
  Internet Address 10.1.2.1/24, Area 0, SID 0, Strict-SPF SID 0
  Label stack Primary label 0 Backup label 0 SRTE label 0
  Process ID 1, Router ID 1.1.1.1, Network Type BROADCAST, Cost: 1
  Transmit Delay is 1 sec, State BDR, Priority 1, MTU 1500, MaxPktSz 1500
  Forward reference No, Unnumbered no,  Bandwidth 1000000 
  RIB LC sync Yes
  Designated Router (ID) 2.2.2.2, Interface address 10.1.2.2
  Backup Designated router (ID) 1.1.1.1, Interface address 10.1.2.1
  Timer intervals configured, Hello 10, Dead 40, Wait 40, Retransmit 5
    Hello due in 00:00:08:985

It is Network Type BROADCAST, the DR is R2 (2.2.2.2), R1 itself is the BDR, and the timers are Hello 10, Dead 40. The LSDB holds a Network-LSA for this segment (Link ID 10.1.2.2).

STEP 0 R1 LSDB
RP/0/RP0/CPU0:R1#show ospf database
Sun Sep  6 05:02:42.895 UTC


            OSPF Router with ID (1.1.1.1) (Process ID 1)

		Router Link States (Area 0)

Link ID         ADV Router      Age         Seq#       Checksum Link count
1.1.1.1         1.1.1.1         42          0x80000003 0x00b343 2
2.2.2.2         2.2.2.2         36          0x80000003 0x00ded7 3
3.3.3.3         3.3.3.3         36          0x80000003 0x0004d3 2

		Net Link States (Area 0)

Link ID         ADV Router      Age         Seq#       Checksum
10.1.2.2        2.2.2.2         43          0x80000001 0x0031e5
10.2.3.3        3.3.3.3         37          0x80000001 0x0046c1

Changing to point-to-point (STEP 1)

Configure network point-to-point on both R1 and R2.

STEP 1 configuration applied to R1 and R2
router ospf 1
 area 0
  interface GigabitEthernet0/0/0/0
   network point-to-point
  !
 !
!
STEP 1 R1 point-to-point
GigabitEthernet0/0/0/0 is up, line protocol is up 
  Internet Address 10.1.2.1/24, Area 0, SID 0, Strict-SPF SID 0
  Label stack Primary label 0 Backup label 0 SRTE label 0
  Process ID 1, Router ID 1.1.1.1, Network Type POINT_TO_POINT, Cost: 1
  Transmit Delay is 1 sec, State POINT_TO_POINT, MTU 1500, MaxPktSz 1500
  Forward reference No, Unnumbered no,  Bandwidth 1000000 
  RIB LC sync Yes
  Timer intervals configured, Hello 10, Dead 40, Wait 40, Retransmit 5
    Hello due in 00:00:06:819
  Index 1/1, flood queue length 0

The DR and BDR lines are gone entirely and State is now POINT_TO_POINT. The Hello and dead intervals stay at 10/40. In show ospf neighbor the role after the / becomes -.

STEP 1 R1 neighbors
RP/0/RP0/CPU0:R1#show ospf neighbor
Sun Sep  6 05:06:30.673 UTC

* Indicates MADJ interface
# Indicates Neighbor awaiting BFD session up

Neighbors for OSPF 1

Neighbor ID     Pri   State           Dead Time   Address         Interface
2.2.2.2         1     FULL/  -        00:00:35    10.1.2.2        GigabitEthernet0/0/0/0
    Neighbor is up for 00:01:30

Total neighbor count: 1

The Network-LSA for this segment (10.1.2.2) has disappeared from the LSDB; the only one left is 10.2.3.3 for the R2–R3 segment. In exchange, R1’s Router-LSA has grown from two links to three.

STEP 1 R1 LSDB (one Network-LSA fewer)
RP/0/RP0/CPU0:R1#show ospf database
Sun Sep  6 05:06:32.582 UTC


            OSPF Router with ID (1.1.1.1) (Process ID 1)

		Router Link States (Area 0)

Link ID         ADV Router      Age         Seq#       Checksum Link count
1.1.1.1         1.1.1.1         92          0x80000005 0x006777 3
2.2.2.2         2.2.2.2         93          0x80000005 0x000a98 4
3.3.3.3         3.3.3.3         263         0x80000003 0x0004d3 2

		Net Link States (Area 0)

Link ID         ADV Router      Age         Seq#       Checksum
10.2.3.3        3.3.3.3         264         0x80000001 0x0046c1

The extra link is there because a point-to-point link is advertised as two things: a link to the peer router and a stub network for the segment itself. With no DR there is no Network-LSA, so the segment is represented as a stub.

STEP 1 R1 Router-LSA
RP/0/RP0/CPU0:R1#show ospf database router
Sun Sep  6 05:06:33.580 UTC


            OSPF Router with ID (1.1.1.1) (Process ID 1)

		Router Link States (Area 0)

  LS age: 93
  Options: (No TOS-capability, DC)
  LS Type: Router Links
  Link State ID: 1.1.1.1
  Advertising Router: 1.1.1.1
  LS Seq Number: 80000005
  Checksum: 0x6777
  Length: 60
   Number of Links: 3

    Link connected to: a Stub Network
     (Link ID) Network/subnet number: 1.1.1.1
     (Link Data) Network Mask: 255.255.255.255
      Number of TOS metrics: 0
       TOS 0 Metrics: 1

    Link connected to: another Router (point-to-point)
     (Link ID) Neighboring Router ID: 2.2.2.2
     (Link Data) Router Interface address: 10.1.2.1
      Number of TOS metrics: 0
       TOS 0 Metrics: 1

The difference is visible on the wire too. With the Hellos from before and after the change in a single capture, the DR/BDR fields go from real addresses to 0.0.0.0.

No.TimeSourceDestinationInfo
10.00010.1.2.2224.0.0.5Hello Packet
22.55810.1.2.1224.0.0.5Hello Packet
39.48410.1.2.2224.0.0.5Hello Packet
411.87910.1.2.1224.0.0.5Hello Packet
516.01110.1.2.1224.0.0.5Hello Packet
619.47510.1.2.2224.0.0.5Hello Packet
725.62310.1.2.1224.0.0.5Hello Packet
1036.18110.1.2.2224.0.0.5Hello Packet
1145.23110.1.2.1224.0.0.5Hello Packet
1245.26110.1.2.210.1.2.1DB Description

Up to No.4 these are broadcast Hellos; from No.5 they are point-to-point Hellos. R2 switched a little later, so No.6 and No.8 still carry a DR. Once both sides show 0.0.0.0, the DBD exchange starts at No.12 and the adjacency is rebuilt.

No.4 Hello (broadcast, before the change)
Open Shortest Path First
    OSPF Header
        Version: 2
        Message Type: Hello Packet (1)
        Packet Length: 48
        Source OSPF Router: 1.1.1.1
        Area ID: 0.0.0.0 (Backbone)
        Checksum: 0xce8f [correct]
        Instance ID: Base IPv4 Unicast Instance (0)
        Auth Type: Null (0)
        Auth Data (none): 0000000000000000
    OSPF Hello Packet
        Network Mask: 255.255.255.0
        Hello Interval [sec]: 10
        Options: 0x12, (L) LLS Data block, (E) External Routing
            0... .... = DN: Not set
            .0.. .... = (O) Opaque: Not set
            ..0. .... = (DC) Demand Circuits: Not supported
            ...1 .... = (L) LLS Data block: Present
            .... 0... = (N) NSSA: Not supported
            .... .0.. = (MC) Multicast: Not capable
            .... ..1. = (E) External Routing: Capable
            .... ...0 = (MT) Multi-Topology Routing: No
        Router Priority: 1
        Router Dead Interval [sec]: 40
        Designated Router: 10.1.2.2
        Backup Designated Router: 10.1.2.1
        Active Neighbor: 2.2.2.2
    OSPF LLS Data Block
        Checksum: 0x7fc4
        LLS Data Length: 32 bytes
        Extended options TLV
            TLV Type: 1
            TLV Length: 4
            Options: 0x00000001, (LR) LSDB Resynchronization
                .... .... .... .... .... .... .... ..0. = (RS) Restart Signal: Not set
                .... .... .... .... .... .... .... ...1 = (LR) LSDB Resynchronization: Set
        Local Interface ID
            TLV Type: 18
            TLV Length: 4
            Local Interface ID: 00000003
        Unknown LLS TLV
            TLV Type: 32768
            TLV Length: 8
Download the pcap of the packet in the tshark output above (No.4 Hello)
No.5 Hello (point-to-point, after the change)
Open Shortest Path First
    OSPF Header
        Version: 2
        Message Type: Hello Packet (1)
        Packet Length: 44
        Source OSPF Router: 1.1.1.1
        Area ID: 0.0.0.0 (Backbone)
        Checksum: 0xea9c [correct]
        Instance ID: Base IPv4 Unicast Instance (0)
        Auth Type: Null (0)
        Auth Data (none): 0000000000000000
    OSPF Hello Packet
        Network Mask: 255.255.255.0
        Hello Interval [sec]: 10
        Options: 0x12, (L) LLS Data block, (E) External Routing
            0... .... = DN: Not set
            .0.. .... = (O) Opaque: Not set
            ..0. .... = (DC) Demand Circuits: Not supported
            ...1 .... = (L) LLS Data block: Present
            .... 0... = (N) NSSA: Not supported
            .... .0.. = (MC) Multicast: Not capable
            .... ..1. = (E) External Routing: Capable
            .... ...0 = (MT) Multi-Topology Routing: No
        Router Priority: 1
        Router Dead Interval [sec]: 40
        Designated Router: 0.0.0.0
        Backup Designated Router: 0.0.0.0
    OSPF LLS Data Block
        Checksum: 0x7fc4
        LLS Data Length: 32 bytes
        Extended options TLV
            TLV Type: 1
            TLV Length: 4
            Options: 0x00000001, (LR) LSDB Resynchronization
                .... .... .... .... .... .... .... ..0. = (RS) Restart Signal: Not set
                .... .... .... .... .... .... .... ...1 = (LR) LSDB Resynchronization: Set
        Local Interface ID
            TLV Type: 18
            TLV Length: 4
            Local Interface ID: 00000003
        Unknown LLS TLV
            TLV Type: 32768
            TLV Length: 8
Download the pcap of the packet in the tshark output above (No.5 Hello)

Mismatched network types prevent an adjacency (STEP 2 and 3)

Put only R1 back to the default (broadcast) and leave R2 as point-to-point. Both ends still have Hello/dead 10/40, so a timer mismatch cannot explain what happens.

R1 becomes the DR on its own, and the neighbor never gets past INIT.

STEP 2 R1 the broadcast side
GigabitEthernet0/0/0/0 is up, line protocol is up 
  Internet Address 10.1.2.1/24, Area 0, SID 0, Strict-SPF SID 0
  Label stack Primary label 0 Backup label 0 SRTE label 0
  Process ID 1, Router ID 1.1.1.1, Network Type BROADCAST, Cost: 1
  Transmit Delay is 1 sec, State DR, Priority 1, MTU 1500, MaxPktSz 1500
  Forward reference No, Unnumbered no,  Bandwidth 1000000 
  RIB LC sync Yes
  Designated Router (ID) 1.1.1.1, Interface address 10.1.2.1
  No backup designated router on this network
  Timer intervals configured, Hello 10, Dead 40, Wait 40, Retransmit 5
    Hello due in 00:00:08:082
  Index 1/1, flood queue length 0
  Next 0(0)/0(0)
STEP 2 R1 neighbors
RP/0/RP0/CPU0:R1#show ospf neighbor
Sun Sep  6 05:10:54.370 UTC

* Indicates MADJ interface
# Indicates Neighbor awaiting BFD session up

Neighbors for OSPF 1

Neighbor ID     Pri   State           Dead Time   Address         Interface
2.2.2.2         1     INIT/DROTHER    00:00:37    10.1.2.2        GigabitEthernet0/0/0/0

Total neighbor count: 1

R2 does not count it as a neighbor at all.

STEP 2 R2 the point-to-point side
GigabitEthernet0/0/0/0 is up, line protocol is up 
  Internet Address 10.1.2.2/24, Area 0, SID 0, Strict-SPF SID 0
  Label stack Primary label 0 Backup label 0 SRTE label 0
  Process ID 1, Router ID 2.2.2.2, Network Type POINT_TO_POINT, Cost: 1
  Transmit Delay is 1 sec, State POINT_TO_POINT, MTU 1500, MaxPktSz 1500
  Forward reference No, Unnumbered no,  Bandwidth 1000000 
  RIB LC sync Yes
  Timer intervals configured, Hello 10, Dead 40, Wait 40, Retransmit 5
    Hello due in 00:00:09:320
  Index 1/1, flood queue length 0
  Next 0(0)/0(0)
  Last flood scan length is 1, maximum is 2
  Last flood scan time is 0 msec, maximum is 0 msec
  LS Ack List: current length 0, high water mark 2
  Neighbor Count is 0, Adjacent neighbor count is 0

The reason is recorded in show ospf trace hello. IOS XR keeps Hello processing in a trace buffer, so it can be read back later without having set up a debug.

STEP 2 R2 why the Hello is discarded (excerpt from show ospf trace hello)
405  Sep  6 05:11:29.430  ospf_rcv_hello: intf Gi0/0/0/0 area 0.0.0.0 from 1.1.1.1 10.1.2.1
406  Sep  6 05:11:29.430  ospf_validate_hello: non-zero DR 10.1.2.1 on non-multiaccess interface, ignoring hello
407  Sep  6 05:11:29.430  ospf_rcv_hello: WARN hello validation failed, ospf_rtr_id 1.1.1.1

A point-to-point interface discards a Hello whose DR field is not zero. The broadcast side made itself the DR and puts its own address in the DR field, and the point-to-point side throws it away.

As a result R1 is left with no OSPF routes at all.

STEP 2 R1 routes
RP/0/RP0/CPU0:R1#show route ospf
Sun Sep  6 05:10:51.396 UTC

% No matching routes found

Putting both ends back to the default in STEP 3 restores the adjacency and the routes. The network type has to be the same on both ends of a link.

STEP 3 R1 after recovery
RP/0/RP0/CPU0:R1#show route ospf
Sun Sep  6 05:13:59.016 UTC

O    2.2.2.2/32 [110/2] via 10.1.2.2, 00:01:22, GigabitEthernet0/0/0/0
O    3.3.3.3/32 [110/3] via 10.1.2.2, 00:01:22, GigabitEthernet0/0/0/0
O    10.2.3.0/24 [110/2] via 10.1.2.2, 00:01:22, GigabitEthernet0/0/0/0

Changing to non-broadcast (NBMA) (STEP 4 and 5)

Configure network non-broadcast on both ends.

STEP 4 configuration applied to R1 and R2
router ospf 1
 area 0
  interface GigabitEthernet0/0/0/0
   network non-broadcast
  !
 !
!
STEP 4 R1 non-broadcast (no neighbor configured)
GigabitEthernet0/0/0/0 is up, line protocol is up 
  Internet Address 10.1.2.1/24, Area 0, SID 0, Strict-SPF SID 0
  Label stack Primary label 0 Backup label 0 SRTE label 0
  Process ID 1, Router ID 1.1.1.1, Network Type NON_BROADCAST, Cost: 1
  Transmit Delay is 1 sec, State DR, Priority 1, MTU 1500, MaxPktSz 1500
  Forward reference No, Unnumbered no,  Bandwidth 1000000 
  RIB LC sync Yes
  Designated Router (ID) 1.1.1.1, Interface address 10.1.2.1
  No backup designated router on this network
  Timer intervals configured, Hello 30, Dead 120, Wait 120, Retransmit 5
    Hello due in 00:00:06:525
  Index 1/1, flood queue length 0
  Next 0(0)/0(0)
  Last flood scan length is 1, maximum is 1
  Last flood scan time is 0 msec, maximum is 0 msec
  LS Ack List: current length 0, high water mark 3
  Neighbor Count is 0, Adjacent neighbor count is 0

The timers change to Hello 30, Dead 120, Wait 120, a DR is still elected (R1 made itself the DR), but not a single neighbor is found, because non-broadcast does not look for neighbors with multicast.

The capture shows that not one OSPF packet crosses the link during this time. With no configured peer to unicast to, not even Hellos are sent.

No.TimeSourceDestinationInfo
10.00010.1.2.1224.0.0.5Hello Packet
22.21410.1.2.2224.0.0.5Hello Packet
39.82810.1.2.1224.0.0.5Hello Packet
412.06010.1.2.2224.0.0.5Hello Packet
521.54310.1.2.2224.0.0.5Hello Packet
6329.59410.1.2.110.1.2.2Hello Packet
7341.45510.1.2.210.1.2.1Hello Packet
8341.46010.1.2.110.1.2.2DB Description
23358.81410.1.2.110.1.2.2Hello Packet
24370.47310.1.2.210.1.2.1Hello Packet

There is nothing at all between No.5 (21.5 s) and No.6 (329.6 s), about five minutes. No.6 is the first Hello after neighbor was configured, and its destination is the peer’s interface address 10.1.2.2 rather than 224.0.0.5.

STEP 5 configuration applied to R1 and R2
router ospf 1
 area 0
  interface GigabitEthernet0/0/0/0
   neighbor 10.1.2.2     ! neighbor 10.1.2.1 on R2
  !
 !
!
No.6 Hello (non-broadcast, sent as unicast)
Open Shortest Path First
    OSPF Header
        Version: 2
        Message Type: Hello Packet (1)
        Packet Length: 44
        Source OSPF Router: 1.1.1.1
        Area ID: 0.0.0.0 (Backbone)
        Checksum: 0xde36 [correct]
        Instance ID: Base IPv4 Unicast Instance (0)
        Auth Type: Null (0)
        Auth Data (none): 0000000000000000
    OSPF Hello Packet
        Network Mask: 255.255.255.0
        Hello Interval [sec]: 30
        Options: 0x12, (L) LLS Data block, (E) External Routing
            0... .... = DN: Not set
            .0.. .... = (O) Opaque: Not set
            ..0. .... = (DC) Demand Circuits: Not supported
            ...1 .... = (L) LLS Data block: Present
            .... 0... = (N) NSSA: Not supported
            .... .0.. = (MC) Multicast: Not capable
            .... ..1. = (E) External Routing: Capable
            .... ...0 = (MT) Multi-Topology Routing: No
        Router Priority: 1
        Router Dead Interval [sec]: 120
        Designated Router: 10.1.2.1
        Backup Designated Router: 0.0.0.0
    OSPF LLS Data Block
        Checksum: 0x7fc4
        LLS Data Length: 32 bytes
        Extended options TLV
            TLV Type: 1
            TLV Length: 4
            Options: 0x00000001, (LR) LSDB Resynchronization
                .... .... .... .... .... .... .... ..0. = (RS) Restart Signal: Not set
                .... .... .... .... .... .... .... ...1 = (LR) LSDB Resynchronization: Set
        Local Interface ID
            TLV Type: 18
            TLV Length: 4
            Local Interface ID: 00000003
        Unknown LLS TLV
            TLV Type: 32768
            TLV Length: 8
Download the pcap of the packet in the tshark output above (No.6 Hello)

Once neighbor is in place the adjacency forms and a DR/BDR is elected.

STEP 5 R1 after configuring neighbors statically
GigabitEthernet0/0/0/0 is up, line protocol is up 
  Internet Address 10.1.2.1/24, Area 0, SID 0, Strict-SPF SID 0
  Label stack Primary label 0 Backup label 0 SRTE label 0
  Process ID 1, Router ID 1.1.1.1, Network Type NON_BROADCAST, Cost: 1
  Transmit Delay is 1 sec, State BDR, Priority 1, MTU 1500, MaxPktSz 1500
  Forward reference No, Unnumbered no,  Bandwidth 1000000 
  RIB LC sync Yes
  Designated Router (ID) 2.2.2.2, Interface address 10.1.2.2
  Backup Designated router (ID) 1.1.1.1, Interface address 10.1.2.1
  Timer intervals configured, Hello 30, Dead 120, Wait 120, Retransmit 5
    Hello due in 00:00:21:917
  Index 1/1, flood queue length 0
  Next 0(0)/0(0)
  Last flood scan length is 1, maximum is 1
  Last flood scan time is 0 msec, maximum is 0 msec
  LS Ack List: current length 0, high water mark 3
  Neighbor Count is 1, Adjacent neighbor count is 1

Changing to point-to-multipoint (STEP 6)

Remove neighbor and configure network point-to-multipoint on both ends.

STEP 6 configuration applied to R1 and R2
router ospf 1
 area 0
  interface GigabitEthernet0/0/0/0
   no neighbor 10.1.2.2  ! no neighbor 10.1.2.1 on R2
   network point-to-multipoint
  !
 !
!
STEP 6 R1 point-to-multipoint
GigabitEthernet0/0/0/0 is up, line protocol is up 
  Internet Address 10.1.2.1/24, Area 0, SID 0, Strict-SPF SID 0
  Label stack Primary label 0 Backup label 0 SRTE label 0
  Process ID 1, Router ID 1.1.1.1, Network Type POINT_TO_MULTIPOINT, Cost: 1
  Transmit Delay is 1 sec, State POINT_TO_MULTIPOINT, MTU 1500, MaxPktSz 1500
  Forward reference No, Unnumbered no,  Bandwidth 1000000 
  RIB LC sync Yes
  Timer intervals configured, Hello 30, Dead 120, Wait 120, Retransmit 5
    Hello due in 00:00:02:902
  Index 1/1, flood queue length 0

The timers are the same 30/120 as non-broadcast, but no DR/BDR is elected and the adjacency forms without any neighbor configuration. Hellos go to the multicast address (224.0.0.5).

No.4 Hello (point-to-multipoint)
Open Shortest Path First
    OSPF Header
        Version: 2
        Message Type: Hello Packet (1)
        Packet Length: 48
        Source OSPF Router: 1.1.1.1
        Area ID: 0.0.0.0 (Backbone)
        Checksum: 0xe630 [correct]
        Instance ID: Base IPv4 Unicast Instance (0)
        Auth Type: Null (0)
        Auth Data (none): 0000000000000000
    OSPF Hello Packet
        Network Mask: 255.255.255.0
        Hello Interval [sec]: 30
        Options: 0x12, (L) LLS Data block, (E) External Routing
            0... .... = DN: Not set
            .0.. .... = (O) Opaque: Not set
            ..0. .... = (DC) Demand Circuits: Not supported
            ...1 .... = (L) LLS Data block: Present
            .... 0... = (N) NSSA: Not supported
            .... .0.. = (MC) Multicast: Not capable
            .... ..1. = (E) External Routing: Capable
            .... ...0 = (MT) Multi-Topology Routing: No
        Router Priority: 1
        Router Dead Interval [sec]: 120
        Designated Router: 0.0.0.0
        Backup Designated Router: 0.0.0.0
        Active Neighbor: 2.2.2.2
    OSPF LLS Data Block
        Checksum: 0x7fc4
        LLS Data Length: 32 bytes
        Extended options TLV
            TLV Type: 1
            TLV Length: 4
            Options: 0x00000001, (LR) LSDB Resynchronization
                .... .... .... .... .... .... .... ..0. = (RS) Restart Signal: Not set
                .... .... .... .... .... .... .... ...1 = (LR) LSDB Resynchronization: Set
        Local Interface ID
            TLV Type: 18
            TLV Length: 4
            Local Interface ID: 00000003
        Unknown LLS TLV
            TLV Type: 32768
            TLV Length: 8
Download the pcap of the packet in the tshark output above (No.4 Hello)

What is distinctive about this type is that the segment is advertised as a /32 host route per peer rather than as a subnet. R1’s routing table now has 10.1.2.2/32, which none of the other types produce.

STEP 6 R1 routes (a /32 host route appears)
RP/0/RP0/CPU0:R1#show route ospf
Sun Sep  6 05:30:31.183 UTC

O    2.2.2.2/32 [110/2] via 10.1.2.2, 00:03:17, GigabitEthernet0/0/0/0
O    3.3.3.3/32 [110/3] via 10.1.2.2, 00:03:17, GigabitEthernet0/0/0/0
O    10.1.2.2/32 [110/1] via 10.1.2.2, 00:03:17, GigabitEthernet0/0/0/0
O    10.2.3.0/24 [110/2] via 10.1.2.2, 00:03:17, GigabitEthernet0/0/0/0

In the Router-LSA, alongside the point-to-point link to the peer, R1’s own interface address 10.1.2.1/32 is advertised as a stub with a metric of 0. In a partial mesh, which pairs can reach each other directly differs from pair to pair, so the advertisement is done per pair rather than per subnet.

STEP 6 R1 Router-LSA
RP/0/RP0/CPU0:R1#show ospf database router
Sun Sep  6 05:30:36.515 UTC


            OSPF Router with ID (1.1.1.1) (Process ID 1)

		Router Link States (Area 0)

  LS age: 203
  Options: (No TOS-capability, DC)
  LS Type: Router Links
  Link State ID: 1.1.1.1
  Advertising Router: 1.1.1.1
  LS Seq Number: 8000000e
  Checksum: 0x4d88
  Length: 60
   Number of Links: 3

    Link connected to: a Stub Network
     (Link ID) Network/subnet number: 1.1.1.1
     (Link Data) Network Mask: 255.255.255.255
      Number of TOS metrics: 0
       TOS 0 Metrics: 1

    Link connected to: another Router (point-to-point)
     (Link ID) Neighboring Router ID: 2.2.2.2
     (Link Data) Router Interface address: 10.1.2.1
      Number of TOS metrics: 0
       TOS 0 Metrics: 1

    Link connected to: a Stub Network
     (Link ID) Network/subnet number: 10.1.2.1
     (Link Data) Network Mask: 255.255.255.255
      Number of TOS metrics: 0
       TOS 0 Metrics: 0

Back to the default (STEP 7)

Removing the network configuration returns the interface to broadcast, the default on Ethernet.

STEP 7 R1 restored to the default
GigabitEthernet0/0/0/0 is up, line protocol is up 
  Internet Address 10.1.2.1/24, Area 0, SID 0, Strict-SPF SID 0
  Label stack Primary label 0 Backup label 0 SRTE label 0
  Process ID 1, Router ID 1.1.1.1, Network Type BROADCAST, Cost: 1
  Transmit Delay is 1 sec, State BDR, Priority 1, MTU 1500, MaxPktSz 1500
  Forward reference No, Unnumbered no,  Bandwidth 1000000 
  RIB LC sync Yes
  Designated Router (ID) 2.2.2.2, Interface address 10.1.2.2
  Backup Designated router (ID) 1.1.1.1, Interface address 10.1.2.1
  Timer intervals configured, Hello 10, Dead 40, Wait 40, Retransmit 5
    Hello due in 00:00:03:925
IOS XR also accepts network point-to-multipoint non-broadcast. However show ospf interface reports it as plain POINT_TO_MULTIPOINT, exactly like the ordinary point-to-multipoint type, so the two cannot be told apart there. You have to look at the running-config to know which one is in effect (the output taken during the test is under “Additional” for STEP 7 in the verification section at the end of the article).

Comparing the Hello packets

Putting the Hellos actually seen for the four types side by side, the differences come down to three things: the destination address, the Hello/dead intervals and the DR/BDR fields.

TypeDestinationHello / DeadDR fieldCapture
Broadcast224.0.0.510 / 4010.1.2.2 (the DR’s address)ospf-network-type-p2p.pcap No.4
Point-to-point224.0.0.510 / 400.0.0.0ospf-network-type-p2p.pcap No.5
Non-broadcast10.1.2.2 (unicast)30 / 12010.1.2.1 (the DR’s address)ospf-network-type-nbma.pcap No.6
Point-to-multipoint224.0.0.530 / 1200.0.0.0ospf-network-type-p2mp.pcap No.4

Choosing a type in a design

  • An Ethernet link with only two routers wastes less if it is set to point-to-point. The DR/BDR election (with its 40-second Wait timer) is no longer needed so the adjacency comes up faster, and one Network-LSA fewer keeps the LSDB smaller. As STEP 1 showed, the routes themselves do not change.
  • Always use the same type on both ends. As in STEP 2, even with matching Hello/dead intervals a difference in type alone prevents the adjacency, and because show ospf neighbor still lists the peer in INIT, the failure is hard to diagnose.
  • With non-broadcast, a missing neighbor statement means a complete outage. Nothing at all is sent until it is configured, so a state where only one side is configured deserves attention too.
  • Broadcast and non-broadcast elect a DR/BDR. The details of that election on a segment with many routers are covered in DR and BDR.

Verification config and show output

At every STEP the following three files were collected from all three routers, split per router. The verification config is the ..._run.txt file (the final state is the one from STEP 7).

FileContents
..._show.txtshow version / show interface description / show route / show route ospf / show ospf / show ospf interface / show ospf interface brief / show ospf neighbor / show ospf neighbor detail / show ospf database / show ospf database router / show ospf database network / show ospf statistics interface / show ospf routes / show ospf trace events / show ospf trace adj / show ospf trace hello
..._log.txtshow logging narrowed down to that STEP. A marker is written with logmsg at the start of each STEP and its timestamp is passed to show logging start (STEP 0 holds the whole history since boot)
..._run.txtshow running-config at that STEP (that is, the verification config of that STEP)

Several STEPs also include the record of the commands checked on R1 right after the change (..._check.txt).

STEP 0: initial state (the default, broadcast) — BROADCAST, Hello 10/Dead 40, a DR/BDR is elected

Routershowsyslogrunning-config
R1showlogrun
R2showlogrun
R3showlogrun

STEP 1: network point-to-point on both R1 and R2 — POINT_TO_POINT, no DR/BDR, the Network-LSA disappears

Routershowsyslogrunning-config
R1showlogrun
R2showlogrun
R3showlogrun

STEP 2: put only R1 back to the default (mismatched types) — no adjacency forms and every OSPF route is lost

Routershowsyslogrunning-config
R1showlogrun
R2showlogrun
R3showlogrun

Additional: R1 the mismatch

STEP 3: put R2 back to the default as well — recovered

Routershowsyslogrunning-config
R1showlogrun
R2showlogrun
R3showlogrun

STEP 4: network non-broadcast on both ends — NON_BROADCAST, Hello 30/Dead 120, no neighbor is discovered

Routershowsyslogrunning-config
R1showlogrun
R2showlogrun
R3showlogrun

Additional: R1 NBMA without neighbors

STEP 5: configure neighbor on both ends — the adjacency forms and Hellos become unicast

Routershowsyslogrunning-config
R1showlogrun
R2showlogrun
R3showlogrun

Additional: R1 NBMA with static neighbors

STEP 6: remove neighbor and configure network point-to-multipoint — POINT_TO_MULTIPOINT, no DR, no neighbor needed

Routershowsyslogrunning-config
R1showlogrun
R2showlogrun
R3showlogrun

Additional: R1 point-to-multipoint

STEP 7: remove the network configuration (final state) — restored to the same setup as the initial state

Routershowsyslogrunning-config
R1showlogrun
R2showlogrun
R3showlogrun

Additional: R1 back to the default / R1 point-to-multipoint non-broadcast

The three capture files are below. All were taken on the R1–R2 link.

Download the capture of the switch to point-to-point, STEP 1 (ospf-network-type-p2p.pcap)

Download the capture of non-broadcast, STEP 4-5 (ospf-network-type-nbma.pcap)

Download the capture of point-to-multipoint, STEP 6 (ospf-network-type-p2mp.pcap)

References

RFCTitleSummary
RFC 2328OSPF Version 2The definition of the network types (Section 1.2) and the per-type interface behaviour (Section 9).

Related articles