Expand all | Collapse all

Default routes for parallel links

Jump to Best Answer
  • 1.  Default routes for parallel links

    Posted 04-19-2019 07:39

    I have two EX4300 VCs that will be connected via two point to point wireless links. I have OSPF running, and a default route on VC-B pointing to VC-A, (VC-A has a default route to the edge firewall). Internal routing with OSPF is working as expected, having taken an old fiber connection that was the previous path out of the config, but the static route on VC-B is still using the fiber cable. I am trying to take advantage of both link when routing default traffic with failover in case one goes down.

    Can I just add both paths as next-hops in the static route config on VC-B, and have them multipath?

  • 2.  RE: Default routes for parallel links
    Best Answer

    Posted 04-19-2019 10:52



    To your question: Yes, you can add two next-hops to allow the switch to send different flows per link. but in order to get traffic load balanced you will need to configure a policy for that and apply it to the routing-options forwarding-table. for example:


    routing-options {
        static {
            route next-hop [ ];
            route next-hop [ ];
        forwarding-table {
            export load-balance;


    The load balancing policy for the switch to actually use both links to send traffic can be as follows, this will match any prefix:


    policy-options {
        policy-statement load-balance {
            then {
                load-balance per-packet;


    Then just check that both next hops are installed by issuing:


    user@switch> show route forwarding-table destination


    Some documentation to add:



    Configuring Per-Packet Load Balancing



  • 3.  RE: Default routes for parallel links

    Posted 04-20-2019 08:39

    Thank you!
    The forwarding tables for both VCs are now showing all the required interfaces. This was excatly what I was looking for.

  • 4.  RE: Default routes for parallel links

    Posted 04-22-2019 10:21

    Happy to know that. 🙂


    Best Regards