Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
ON THIS PAGE
 

Failover Options

Juniper Mist Edge delivers multiple layers of failover and redundancy. The design ensures network survivability and continuous connectivity for wireless clients. The deployment includes clustering, dual tunnels, upstream resource monitoring, and auto-preemption.

Intra–Cluster Failover

Deploy Mist Edges in an active–active cluster that load-balances APs across all members. The configuration ensures high availability and zero idle capacity. If a Mist Edge in a cluster becomes unavailable, APs that tunnel to it fail over to another active Mist Edge within the same Layer 2 (L2) cluster. Design L2 redundancy by provisioning Mist Edges to operate at 80% capacity . This practice ensures enough headroom for failover.

Inter–Cluster Failover

Configure multiple Mist Edge clusters for L3 redundancy.

  • Primary–Secondary Clusters: Configure APs to prefer a primary cluster and fail over to a secondary cluster if the primary becomes unreachable.

  • Dual Tunnel: Configure APs to tunnel traffic to two different Mist Edge clusters that are geographically or L3 separated.

  • Maximize Resource Utilization: Configure multiple tunnels from the WLAN on APs to balance loads across data centers. A single Mist Edge cluster can act as the primary for one set of tunnels and serve as the secondary for another.

Condition-Based Failover

  • Critical Resource Monitoring (CRM)-Configure CRM to monitor upstream resources using protocols like ARP, PING, or TCP. If a resource becomes unreachable, the Mist Edge temporarily shuts down its tunnel termination service and prompts APs to fail over to the next healthy member. For more information see, Critical Resource Monitoring (CRM).

  • Failover Tunnel Timers-Administrators can fine-tune how quickly an AP fails over by adjusting hello intervals and retry settings for each tunnel. That approach optimizes performance for latency-sensitive applications. For more information see, Failover Tunnel Timers.

Auto-Preemption

Auto-preemption is a cloud-orchestrated process that moves tunnels back to a preferred (primary) Mist Edge cluster after it has recovered.

  • Cloud-based orchestration: A service in the cloud monitors APs failing over and nudges them to switch back to their preferred Mist Edge. That process doesn’t de-authenticate wireless clients.

  • Controlled Timing: Administrators can choose when preemption happens to avoid disrupting the network during peak hours.

    Options include the following:

    • Every 15 minutes: Use this in networks with jittery connectivity to ensure APs move back to the correct cluster. That practice is useful for preventing clients from getting new IP addresses if the secondary cluster has a different IP schema.

    • Time of day: Allows preemption to occur during a specified, typically low-traffic, period.

For more information, see Auto-Preemption.

Anchor Tunnels

When traffic must be tunneled to a DMZ area deeper in data centers, you can use anchor tunnels. Anchor tunnels enable you to configure Juniper Mist Edge to carry all traffic to DMZ and to tunnel specific traffic to another Mist Edge. For more information see, Anchor Tunnel.