SRX

last person joined: yesterday 

Ask questions and share experiences about the SRX Series, vSRX, and cSRX.
  • 1.  LACP / LAG on Chassis Cluster - Interface Monitoring?

    Posted 12-18-2012 05:33

    Hi,

     

    I need some advice to understand something.

     

    When you enable LAG on reth interfaces (with LACP), you specify the minimum-link parameter. So if I have a LAG on one node with two interfaces and specify minimum-link of 1, the reth interface will not go down if only one of the two links fails. If both links fail, it will do a failover to the other node.

     

    Now how does this play together with the interface-monitor statement in the chassis cluster statement? If I put all the interfaces in there with the weight set to 255, will the minimum-link parameter still hold? I guess not? Would I set the monitoring weight to 128 for each LAG member interface in order to only failover when 2 of 2 links go down?

     

    Some config snippets:

     

    Current chassis cluster config:

     

    chassis {
        cluster {
            control-link-recovery;
            reth-count 16;
            network-management {
                cluster-master;
            }
            redundancy-group 0 {
                node 0 priority 100;
                node 1 priority 1;
            }
            redundancy-group 1 {
                node 0 priority 100;
                node 1 priority 1;
                preempt;
                gratuitous-arp-count 5;
                interface-monitor {
                    ge-0/0/3 weight 255;
                    ge-9/0/3 weight 255;
                    ge-2/0/0 weight 255;    
                    ge-11/0/0 weight 255;   
                    ge-2/0/4 weight 255;    
                    ge-11/0/4 weight 255;   
                    ge-2/0/5 weight 255;    
                    ge-11/0/5 weight 255;   
                    ge-2/0/8 weight 255;    
                    ge-11/0/8 weight 255;   
                    ge-2/0/1 weight 255;    
                    ge-11/0/1 weight 255;   
                    ge-11/0/6 weight 255;   
                    ge-2/0/6 weight 255;    
                }                           
            }                               
        }                                   
    }                                  

     

    One of the reth interfaces, consisting of 2 physical ethernet links on each node:

     

    show interfaces reth0 
    vlan-tagging;
    redundant-ether-options {
        redundancy-group 1;
        minimum-links 1;
        lacp {
            passive;
        }
    }

     In this scenarion, if one of the links on one node fails, would it failover (because interface-monitoring for each interface is at 255), or would it stay up because of the minimum-links = 1 statement?

     

    Thanks

     



  • 2.  RE: LACP / LAG on Chassis Cluster - Interface Monitoring?



  • 3.  RE: LACP / LAG on Chassis Cluster - Interface Monitoring?
    Best Answer

    Posted 12-24-2012 12:47

    Hi

     

    Using minimum-link, you only control "how many child interfaces should be up to make reth interface up on primary node" It is not related to failover of RG from node0 to node1. If you really want to use minimum-link knob for RG failover then you have to use 11.2 JUNOS and you will use minimum-link knob with interface monitor of reth interface.

     

    - The minimum-link will bring down the reth interface

    - The interface monitor of that particular reth interface will do the failover of RG from node0 to node1

     

    HTH