Example for Configuring Routing Loop Detection for Routes Imported from BGP to OSPF

This section describes how to configure routing loop detection for routes imported from BGP to OSPF.

Networking Requirements

On the live network, OSPF routes can be imported to a BGP process for redistribution. In such a scenario, routing policies are usually configured on multiple devices to prevent routing loops. If routing policies are incorrectly configured on the devices that import routes, routing loops may occur. To prevent this problem, configure routing loop detection for the routes imported to OSPF.

On the network shown in Figure 1, IBGP peer relationships are established between DeviceA and DeviceB, between DeviceB and DeviceC, between DeviceC and DeviceD, and between DeviceB and DeviceD; an OSPF process is configured on DeviceC and DeviceD. OSPF is configured to import BGP routes on DeviceC, and BGP is configured to import OSPF routes on DeviceD.

Figure 1 Routing loop detection for routes imported from BGP to OSPF

In this example, interface1, interface2, and interface3 represent GE0/1/0, GE0/1/8, and GE0/1/16, respectively.

Configuration Roadmap

The configuration roadmap is as follows:

Disable routing loop detection, which is enabled by default.
Configure IP addresses for interfaces on each device.
Enable OSPF and BGP, and configure basic OSPF and BGP functions.
Configure route import to construct a routing loop.
Check whether a routing loop occurs.
Enable routing loop detection to check whether the routing loop is eliminated.

Procedure

Disable routing loop detection for routes imported to OSPF on each device. This function is enabled by default.
# Disable routing loop detection for routes imported to OSPF. DeviceA is used as an example.
```
<DeviceA> system-view
[~DeviceA] route loop-detect ospf disable
[*DeviceA] commit
```
The configurations of other devices are similar to those of DeviceA. For configuration details, see Configuration Files in this section.
Configure an IP address for each interface.
DeviceA is used as an example.
```
<DeviceA> system-view
[~DeviceA] interface gigabitethernet 0/1/0
[*DeviceA-GigabitEthernet0/1/0] ip address 10.12.1.1 24
[*DeviceA-GigabitEthernet0/1/0] quit
[*DeviceA] commit
```
The configurations of other devices are similar to those of DeviceA. For configuration details, see Configuration Files in this section.

In addition, configure a static route on DeviceA to simulate a looped route.
```
[~DeviceA] ip route-static 10.0.0.0 255.255.255.255 NULL0
[*DeviceA] commit
```

Enable OSPF and BGP, and configure basic OSPF and BGP functions to implement intra-AS communication.

# Enable BGP on DeviceA, and establish an IBGP peer relationship between DeviceA and DeviceB.

[~DeviceA] bgp 100
[*DeviceA-bgp] router-id 1.11.1.1
[*DeviceA-bgp] peer 10.12.1.2 as-number 100
[*DeviceA-bgp] ipv4-family unicast
[*DeviceA-bgp-af-ipv4] peer 10.12.1.2 enable
[*DeviceA-bgp] quit
[*DeviceA] commit

# Enable BGP on DeviceB and establish IBGP peer relationships between DeviceB and DeviceA, between DeviceB and DeviceC, and between DeviceB and DeviceD.

[~DeviceB] bgp 100
[*DeviceB-bgp] router-id 2.22.2.2
[*DeviceB-bgp] peer 10.12.1.1 as-number 100
[*DeviceB-bgp] peer 10.23.1.3 as-number 100
[*DeviceB-bgp] peer 10.24.1.4 as-number 100
[*DeviceB-bgp] ipv4-family unicast
[*DeviceB-bgp-af-ipv4] peer 10.12.1.1 enable
[*DeviceB-bgp-af-ipv4] peer 10.23.1.3 enable
[*DeviceB-bgp-af-ipv4] peer 10.24.1.4 enable
[*DeviceB-bgp-af-ipv4] peer 10.23.1.3 reflect-client
[*DeviceB-bgp] quit
[*DeviceB] commit

# Enable BGP on DeviceC, and establish an IBGP peer relationship between DeviceC and DeviceB.

[~DeviceC] bgp 100
[*DeviceC-bgp] router-id 3.33.3.3
[*DeviceC-bgp] peer 10.23.1.2 as-number 100
[*DeviceC-bgp] ipv4-family unicast
[*DeviceC-bgp-af-ipv4] peer 10.23.1.2 enable
[*DeviceC-bgp] quit
[*DeviceC] commit

# Enable BGP on DeviceD, and establish an IBGP peer relationship between DeviceD and DeviceB.

[~DeviceD] bgp 100
[*DeviceD-bgp] router-id 4.44.4.4
[*DeviceD-bgp] peer 10.24.1.2 as-number 100
[*DeviceD-bgp] ipv4-family unicast
[*DeviceD-bgp-af-ipv4] peer 10.24.1.2 enable
[*DeviceD-bgp] quit
[*DeviceD] commit

# Configure OSPF on DeviceC and DeviceD. DeviceC is used as an example.

[~DeviceC] ospf 1 router-id 3.33.3.3
[*DeviceC-ospf-1] area 0
[*DeviceC-ospf-1-area-0.0.0.0] network 10.34.1.0 0.0.0.255
[*DeviceC-ospf-1-area-0.0.0.0] quit
[*DeviceC-ospf-1] quit
[*DeviceC] commit

Configure route import.

# Configure OSPF on DeviceC to import BGP routes.

[~DeviceC] ospf 1 router-id 3.33.3.3
[*DeviceC-ospf-1] import-route bgp permit-ibgp
[*DeviceC-ospf-1] quit
[*DeviceC] commit

# Configure BGP on DeviceD to import OSPF routes.

[~DeviceD] bgp 100
[*DeviceD-bgp] ipv4-family unicast
[*DeviceD-bgp-af-ipv4] import-route ospf 1
[*DeviceD-bgp] quit
[*DeviceD] commit

View the routing table on each device to check whether a routing loop occurs.

# Check BGP peer information on DeviceB.

[~DeviceB] display bgp peer
 BGP local router ID : 2.22.2.2
 Local AS number : 100
 Total number of peers : 3                 Peers in established state : 3

  Peer                              V          AS  MsgRcvd  MsgSent  OutQ  Up/Down       State   PrefRcv
  10.12.1.1                         4         100      453      458     0 06:30:47 Established        1
  10.23.1.3                         4         100      452      458     0 06:30:46 Established        0
  10.24.1.4                         4         100      451      457     0 06:29:39 Established        3

# Check OSPF neighbor information on DeviceC.

[~DeviceC] display ospf peer
(M) Indicates MADJ neighbor


          OSPF Process 1 with Router ID 3.33.3.3
                Neighbors

 Area 0.0.0.0 interface 10.34.1.3 (GigabitEthernet0/1/0)'s neighbors
 Router ID: 4.44.4.4             Address: 10.34.1.4
   State: Full           Mode:Nbr is Master     Priority: 1
   DR: 10.34.1.4          BDR: 10.34.1.3          MTU: 0
   Dead timer due in  31  sec
   Retrans timer interval: 5
   Neighbor is up for 06h28m21s
   Neighbor Up Time : 2021-08-27 02:59:32
   Authentication Sequence: [ 0 ]

# Check OSPF neighbor information on DeviceD.

[~DeviceD] display ospf peer
(M) Indicates MADJ neighbor


          OSPF Process 1 with Router ID 4.44.4.4
                Neighbors

 Area 0.0.0.0 interface 10.34.1.4 (GigabitEthernet0/1/8)'s neighbors
 Router ID: 3.33.3.3             Address: 10.34.1.3
   State: Full           Mode:Nbr is Slave      Priority: 1
   DR: 10.34.1.4          BDR: 10.34.1.3          MTU: 0
   Dead timer due in  32  sec
   Retrans timer interval: 5
   Neighbor is up for 06h28m25s
   Neighbor Up Time : 2021-08-27 02:59:32
   Authentication Sequence: [ 0 ]

The preceding command outputs show that BGP peer relationships and OSPF neighbor relationships have been established between the devices.

# Check the BGP routing table on DeviceB.

[~DeviceB] display bgp routing-table 10.0.0.0
 
 BGP local router ID : 2.22.2.2
 Local AS number : 100
 Paths:   2 available, 1 best, 1 select, 0 best-external, 0 add-path
 BGP routing table entry information of 10.0.0.0/32:
 RR-client route.
 From: 10.24.1.4 (4.44.4.4)
 Route Duration: 0d00h00m52s
 Relay IP Nexthop: 10.24.1.4
 Relay IP Out-Interface: GigabitEthernet0/1/0
 Original nexthop: 10.24.1.4
 Qos information : 0x0
 AS-path Nil, origin incomplete, MED 1, localpref 100, pref-val 0, valid, internal, best, select, pre 255
 Advertised to such 3 peers:
    10.23.1.3
    10.24.1.4
    10.12.1.1

 BGP routing table entry information of 10.0.0.0/32:
 From: 10.12.1.1 (1.11.1.1)
 Route Duration: 0d22h53m22s
 Relay IP Nexthop: 10.12.1.1
 Relay IP Out-Interface:GigabitEthernet0/1/8
 Original nexthop: 10.12.1.1
 Qos information : 0x0
 AS-path 10, origin incomplete, MED 0, localpref 100, pref-val 0, valid, internal, pre 255, not preferred for AS-Path
 Not advertised to any peer yet

The preceding command output shows that DeviceB has learned the BGP route distributed by DeviceD.

# Check the BGP routing table of DeviceC.

[~DeviceC] display bgp routing-table 10.0.0.0
 BGP local router ID : 3.33.3.3
 Local AS number : 100
 Paths:   1 available, 1 best, 1 select, 0 best-external, 0 add-path
 BGP routing table entry information of 10.0.0.0/32:
 From: 10.23.1.2 (2.22.2.2)
 Route Duration: 0d07h12m30s
 Relay IP Nexthop: 0.0.0.0
 Relay IP Out-Interface: NULL0
 Original nexthop: 10.12.1.1
 Qos information : 0x0
 AS-path 10, origin incomplete, MED 0, localpref 100, pref-val 0, valid, internal, best, select, pre 255
 Originator: 1.11.1.1
 Cluster list: 2.22.2.2
 Not advertised to any peer yet

The preceding command output shows that DeviceC has learned the BGP route distributed by DeviceB.

# Check the routing table of DeviceD.

[~DeviceD] display ospf routing 10.0.0.0
          OSPF Process 1 with Router ID 4.44.4.4

 Destination    : 10.0.0.0/32
 AdverRouter    : 3.33.3.3                 Tag                 : 1
 Cost           : 1                        Type                : Type2
 NextHop        : 10.34.1.3               Interface           : GigabitEthernet0/1/8
 Priority       : Medium                   Age                 : 01h31m18s

The preceding command output shows that DeviceD has learned the OSPF route distributed by DeviceC.

This means that a routing loop occurs among DeviceB, DeviceC, and DeviceD.

Enable routing loop detection on each device.
# Enable routing loop detection for routes imported to OSPF and BGP. DeviceA is used as an example.
```
[~DeviceA] undo route loop-detect ospf disable
[*DeviceA] route loop-detect bgp enable
[*DeviceA] commit
```
In the case of inter-protocol route import, if a routing protocol with a higher preference detects a routing loop, although this protocol increases the cost of the corresponding route, the cost increase will not render the route inactive. As a result, the routing loop cannot be eliminated. If a routing protocol with a lower preference detects a routing loop and increases the cost of the corresponding route, this route will not be preferred over the originally received route. In this case, the routing loop can be eliminated. OSPF has a higher preference than BGP. Therefore, to eliminate the routing loop, BGP needs to reduce the preference of the corresponding route.

Check whether the routing loop is eliminated.

# Check the BGP routing table on DeviceB.

[~DeviceB] display bgp routing-table 10.0.0.0
 BGP local router ID : 2.22.2.2
 Local AS number : 100
 Paths:   1 available, 1 best, 1 select, 0 best-external, 0 add-path
 BGP routing table entry information of 10.0.0.0/32:
 From: 10.12.1.1 (1.11.1.1)
 Route Duration: 1d00h10m02s
 Relay IP Nexthop: 10.12.1.1
 Relay IP Out-Interface: GigabitEthernet0/1/8
 Original nexthop: 10.12.1.1
 Qos information : 0x0
 AS-path 10, origin incomplete, MED 0, localpref 100, pref-val 0, valid, internal, best, select, pre 255
 Advertised to such 2 peers:
    10.23.1.3
    10.24.1.4

The preceding command output shows that DeviceB has learned the route distributed by DeviceA and that DeviceB no longer prefers the route distributed by DeviceD. This means that the routing loop among DeviceB, DeviceC, and DeviceD is eliminated.

Configuration Files

DeviceA configuration file

#
sysname DeviceA
#
interface GigabitEthernet0/1/0
 undo shutdown
 ip address 10.12.1.1 255.255.255.0
#
bgp 100
 router-id 1.11.1.1
 private-4-byte-as enable
 peer 10.12.1.2 as-number 100
 #
 ipv4-family unicast
  undo synchronization
  import-route static
  peer 10.12.1.2 enable
#
ip route-static 10.0.0.0 255.255.255.255 NULL0
#
return

DeviceB configuration file

#
sysname DeviceB
#
interface GigabitEthernet0/1/8
 undo shutdown
 ip address 10.12.1.2 255.255.255.0
#
interface GigabitEthernet0/1/16
 undo shutdown
 ip address 10.24.1.2 255.255.255.0
#
interface GigabitEthernet0/1/0
 undo shutdown
 ip address 10.23.1.2 255.255.255.0
#
bgp 100
 router-id 2.22.2.2
 private-4-byte-as enable
 peer 10.12.1.1 as-number 100
 peer 10.23.1.3 as-number 100
 peer 10.24.1.4 as-number 100
 #
 ipv4-family unicast
  undo synchronization
  peer 10.12.1.1 enable
  peer 10.23.1.3 enable
  peer 10.23.1.3 reflect-client
  peer 10.24.1.4 enable
  peer 10.24.1.4 reflect-client
#
return

DeviceC configuration file

#
sysname DeviceC
#
interface GigabitEthernet0/1/0
 undo shutdown
 ip address 10.34.1.3 255.255.255.0
#
interface GigabitEthernet0/1/8
 undo shutdown
 ip address 10.23.1.3 255.255.255.0
#
bgp 100
 router-id 3.33.3.3
 private-4-byte-as enable
 peer 10.23.1.2 as-number 100
 #
 ipv4-family unicast
  undo synchronization
  peer 10.23.1.2 enable
#
ospf 1 router-id 3.33.3.3
 import-route bgp permit-ibgp
 opaque-capability enable
 area 0.0.0.0
  network 10.23.1.0 0.0.0.255
  network 10.34.1.0 0.0.0.255
#
return

DeviceD configuration file

#
sysname DeviceD
#
interface GigabitEthernet0/1/8
 undo shutdown
 ip address 10.34.1.4 255.255.255.0
#
interface GigabitEthernet0/1/0
 undo shutdown
 ip address 10.24.1.4 255.255.255.0
#
bgp 100
 router-id 4.44.4.4
 private-4-byte-as enable
 peer 10.24.1.2 as-number 100
 #
 ipv4-family unicast
  undo synchronization
  import-route ospf 1
  peer 10.24.1.2 enable
#
ospf 1 router-id 4.44.4.4
 opaque-capability enable
 area 0.0.0.0
#
return