Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
 

Understanding Fast Failover in an EX Series Virtual Chassis

The Virtual Chassis fast failover feature is a hardware-assisted failover mechanism that automatically reroutes traffic and reduces traffic loss in the event of a link failure or switch failure in a Virtual Chassis. If a link between two members fails, traffic flow between those members must be rerouted quickly so that there is minimal traffic loss.

Fast failover is effective only for Virtual Chassis members configured in ring topologies using identical port types.

Fast failover is supported for the following Virtual Chassis:

  • EX3300 Virtual Chassis

  • EX4200 Virtual Chassis

  • Mixed EX4200 and EX4500 Virtual Chassis

  • Mixed EX4200 and EX4550 Virtual Chassis

  • Mixed EX4200, EX4500, and EX4550 Virtual Chassis

  • EX4500 Virtual Chassis

  • Mixed EX4500 and EX4550 Virtual Chassis

  • EX4550 Virtual Chassis

In any mixed Virtual Chassis, fast failover is supported for all Virtual Chassis ports (VCPs), including VCPs that provide interconnections between different types of member switches.

This topic describes the following:

Supported Topologies for Fast Failover

For fast failover to be effective, the Virtual Chassis members must be configured in a ring topology. The ring topology can be formed by using either dedicated VCPs or user-configured uplink VCPs. Fast failover is supported only in a ring topology that uses identical port types, for example, either a topology that uses all dedicated VCPs or one that uses all uplink VCPs. Fast failover is not supported in a ring topology that includes both dedicated VCPs and uplink VCPs. Fast failover is supported, however, in a Virtual Chassis configuration that consists of multiple rings.

How Fast Failover Works

When fast failover is activated, each VCP is automatically configured with a backup port of the same type (dedicated VCP, SFP uplink VCP, or XFP uplink VCP). If a VCP fails, its backup port is used to send traffic. These backup ports act as standby ports and are not meant for load-balancing traffic or any other purposes.

Fast Failover in a Ring Topology Using Dedicated VCPs

When fast failover is activated in a ring topology that uses dedicated VCPs, each VCP is automatically configured with a backup port of the same type. If a VCP fails, its backup port is used to send traffic. Figure 1 shows normal traffic flow in a ring topology using dedicated VCPs.

Figure 1: Normal Traffic Flow in a Ring Topology Using Dedicated VCPsNormal Traffic Flow in a Ring Topology Using Dedicated VCPs

Figure 2 shows traffic redirected by fast failover.

Figure 2: Traffic Redirected by Fast Failover After a Dedicated VCP Link FailureTraffic Redirected by Fast Failover After a Dedicated VCP Link Failure

When the failed link is restored, the Virtual Chassis reconfigures the topology to the topology's original state.

Fast Failover in a Ring Topology Using Uplink Module VCPs

In a ring topology that uses uplink VCPs, each uplink VCP is automatically configured with a backup uplink VCP. If an uplink VCP fails, its backup port is used to send traffic. Figure 3 shows normal traffic flow in a ring topology using SFP uplink VCPs. Normal traffic flow in a ring topology using XFP uplink VCPs is the same.

Note:

To use SFP or XFP uplink ports as VCPs, you must configure them to be VCPs using the request virtual-chassis vc-port command. Once configured, they are converted into VCPs. For example, xe-0/1/0 becomes vcp-255/1/0 after you configure it to be a VCP.

Figure 4 shows traffic redirected by fast failover.

In a ring topology that uses SFP uplink VCPs, there are four ports per module. Consecutive pairs of ports are automatically configured as backup ports for each other. For example, if a Virtual Chassis member has an SFP uplink module installed, uplink module VCPs ge-0/1/0 and ge-0/1/1 are automatically configured as backup ports for each other. Similarly, ports ge-0/1/2 and ge-0/1/3 are automatically configured as backup ports for each other.

In a ring topology that uses XFP uplink module VCPs, there are only two ports per uplink module. Similarly to a topology that uses SFP uplink module VCPs, each port is automatically configured to back up the other port in the uplink module (for example, xe-0/1/0 is the backup for xe-0/1/1).

Fast Failover in a Virtual Chassis Configuration Using Multiple Ring Topologies

Fast failover is supported in a Virtual Chassis configuration with a multiple-ring topology, as shown in Figure 5.

Figure 5: Traffic Redirected by Fast Failover After VCP Link Failures in a Topology with Multiple RingsTraffic Redirected by Fast Failover After VCP Link Failures in a Topology with Multiple Rings

In this scenario, the Virtual Chassis configuration has three rings: two rings that use dedicated VCPs and one ring that uses SFP uplink module VCPs. Fast failover works independently on each ring. Each dedicated VCP in a ring is backed up by another dedicated VCP. Similarly, each SFP uplink module VCP is backed up by another SFP uplink module VCP. Fast failover does not support a ring topology consisting of a mix of dedicated VCPs and uplink module VCPs.

Effects of Topology Changes on a Fast Failover Configuration

When the fast failover feature is activated, topology changes to the Virtual Chassis configuration do not affect the fast failover configuration. In the event of a link or switch failure, fast failover functions normally.