Configuring Bidirectional Forwarding Detection

Bidirectional Forwarding Detection (BFD) protocol is a mechanism for fast detection of failures of forwarding paths. BFD detects path failures in the order of milliseconds. BFD is used with dynamic routing protocols.

In BFD operation, routing peers exchange BFD packets at a negotiated interval. If a packet is not received from a peer within the negotiated interval plus grace interval, the peer is considered to be dead and a notification will be sent to the set of registered routing protocols. In turn, the routing protocols recalculate the best path and reprogram the routing table.

BFD supports smaller time interval, when compared to the timers provided by the routing protocols, thus resulting in faster detection of failures.

The Citrix ADC appliance supports BFD for the following routing protocols: BGP (IPv4 and IPv6), OSPFv2 (IPv4), and OSPFv3 (IPv6). BFD support in the Citrix ADC appliance is compliant with RFCs 5880, 5881, and 5883.

Points to Consider for Configuring Bidirectional Forwarding Detection

Before you start configuring BFD, consider the following points:

• Make sure that you understand the different components of BFD, described in RFCs 5880, 5881, and 5883.
• BFD on a Citrix ADC appliance is supported for the following routing protocols:
  • BGP (IPv4 and IPv6)
  • OSPFv2 (IPv4)
  • OSPFv3 (IPv6)
• 在Citrix BFD ADC设备不支持the following routing protocols:
  • ISIS
  • RIP (IPv4)
  • RIPng (IPv6)
• The following BFD features are not supported on a Citrix ADC appliance:
  • BFD Echo mode
  • BFD Authentication
  • BFD Demand asynchronous mode
• The minimum values for BFD interval and BFD Rx timers are 100 milliseconds.
• When BFD is used in a topology with shared IP addresses (for example, Layer 2 high availability setup with SNIP addresses or a cluster setup with striped IP addresses), BFD brings down the active sessions during a failover because the BFD failure detection time (order of milliseconds) is lesser than the HA failover detection interval (3–4 seconds). Therefore, Citrix recommends usage of Graceful restart in layer-2 HA topologies as the routes are retained during the failover process.

Configuration Steps

Configuring BFD on a Citrix ADC appliance consists of the following tasks:

• Configure BFD Parameters
• Configure BFD Support for Dynamic Routing Protocols

Configure BFD Parameters

The Citrix ADC appliance provides separate BFD session parameters for:

• single hop sessions
• IPv4 multiple hop sessions
• IPv6 multiple hop sessions

If you do not configure BFD parameters for a type of session, the default values are applied for that session.

The default value of each BFD parameter is same for single hop sessions, IPv4 multiple hop sessions, and IPv6 multiple hop sessions. The following table displays the default value of each BFD parameter.

IMPORTANT:

Mellanox NICs in a Citrix appliance take around 1500 ms to initialize. You must set the BFD timers to more than 1500 ms for a Citrix ADC appliance with Mellanox NICs. Citrix recommends setting the BFD timers to 3000 ms:

• Interval Tx = 600 ms
• Minimum Rx = 600 ms
• Multipler = 5

Configuring BFD Parameters for a Single Hop Session

配置的男朋友D parameters for a single hop session by using theVTYSHcommand line, at the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns# configure terminal ns(config)# interface vlan3 ns(config-if)# bfd singlehop-peer interval 200 minrx 200 multiplier 5 ns(config-if)# exit 

Configuring BFD Parameters for IPv4 Multiple Hop Sessions

配置的男朋友D parameters for IPv4 multiple hop sessions by using theVTYSHcommand line, at the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns #配置终端ns(配置)# bfd multihop-peer 20.20.20.138 interval 300 minrx 300 multiplier 5 ns(config)# exit 

Configuring BFD Parameters for IPv6 Multiple Hop Sessions

配置的男朋友D parameters for IPv6 multiple hop sessions by using theVTYSHcommand line, at the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns(config)# bfd multihop-peer ipv6 20fe:125::138 interval 500 minrx 500 multiplier 5 ns(config)# exit 

Configure BFD Support for Dynamic Routing Protocols

You can enable BFD for a dynamic routing protocol for a type of session with a peer. For example, single hop and multiple hops. The Citrix ADC appliance applies the relevant BFD parameter settings to the session.

Configuring BFD for an IPv4 BGP Single Hop Session

配置的男朋友D for an IPv4 BGP single hop session by using theVTYSHcommand line, at the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns #配置终端ns(配置)#路由器bgp 1 ns(config-router)#neighbor 20.20.20.138 remote-as 1 ns(config-router)#neighbor 20.20.20.138 fall-over bfd ns(config-router)#redistribute kernel ns(config-router)#exit 

Configuring BFD for an IPv4 BGP Multiple Hop Session

配置的男朋友D for an IPv4 BGP multiple hop session by using theVTYSHcommand line, at the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns #配置终端ns(配置)#路由器bgp 1 ns(config-router)#neighbor 20.20.20.138 remote-as 1 ns(config-router)#neighbor 20.20.20.138 fall-over bfd multihop ns(config-router)#redistribute kernel ns(config-router)#exit 

Configuring BFD for an IPv6 BGP Single Hop Session

配置的男朋友D for an IPv6 BGP single hop session by using theVTYSHcommand line, at the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns #配置终端ns(配置)#路由器bgp 1 ns(config-router)#neighbor 30fe:123::124 remote-as 1 ns(config-router)#neighbor 30fe:123::124 fall-over bfd ns(config-router)#address-family ipv6 ns(config-router-af)#neighbor 30fe:123::124 activate ns(config-router-af)#redistribute kernel ns(config-router-af)#exit 

Configuring BFD for an IPv6 BGP Multiple Hop Session

配置的男朋友D for an IPv6 BGP multiple hop session by using theVTYSHcommand line, at the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns #配置终端ns(配置)# bfd multihop-peer ipv6 20fe:125::138 interval 500 minrx 500 multiplier 5 ns(config)#router bgp 1 ns(config-router)#neighbor 20fe:125::138 remote-as 1 ns(config-router)#neighbor 20fe:125::138 fall-over bfd multihop ns(config-router)#address-family ipv6 ns(config-router-af)#neighbor 20fe:125::138 activate ns(config-router-af)#redistribute kernel ns(config-router-af)#end 

Configuring BFD for OSPFv2 (IPv4) on Interfaces

You can enable BFD on all or on a specific interface that uses the OSPFv2 protocol.

配置的男朋友D for OSPFv2 on all interfaces by using theVTYSHcommand line:

At the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns# configure terminal ns(config)#router ospf 1 ns(config-router)#bfd all-interfaces ns(config-router)#redistribute kernel ns(config-router)#exit 

配置的男朋友D for OSPFv2 on a specific interface by using theVTYSHcommand line:

At the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns# configure terminal ns(config)# interface vlan5 ns(config-if)# ip ospf bfd ns(config-if)# exit 

Configuring BFD for OSPFv3 (IPv6) on Interfaces

You can enable BFD on all or on a specific interface that uses the OSPFv3 protocol.

配置的男朋友D for OSPFv3 on all interfaces by using theVTYSHcommand line:

At the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns# configure terminal ns(config)#router ipv6 ospf 10 ns(config-router)#bfd all-interfaces ns(config-router)#redistribute kernel ns(config-router)#exit 

配置的男朋友D for OSPFv3 on a specific interface by using theVTYSHcommand line:

At the command prompt, type the following commands, in the order shown:

Sample configuration:

> vtysh ns# configure terminal ns(config)# interface vlan15 ns(config-if)# ipv6 ospf bfd ns(config-if)# exit