01-30-2013 11:54 AM
I've seen example configurations for both static and dynamic (LACP) groups, but haven't seen anything about when you would use one or the other.
Our core will have two FCX stacks interconnected by multiple 10G links. What would be reasons to choose static or dynamic? What's a good way to balance the traffic?
01-31-2013 02:11 PM
Static LAGs are just a bunch of aggregated links. No signaling takes place between two peers and therefore a misconfiguration on one side can be difficult to troubleshoot.
Dynamic LAGs use LACP packets to establish and negotiate the link which means the chance of misconfiguration is reduced (for example the link won't form if it's only configured on one side).
The important thing to note with LACP is that a node can either be passive (just listening) or active (advertising), but at least one side needs to be active to establish the link.
Brocade(config)#interface ethernet 1/1/1 to 1/1/4
Brocade(config-mif-1/1/1-1/1/4)#link-aggregate configure key 10000
Brocade(config-mif-1/1/1-1/1/4)#interface ethernet 3/1/5 to 3/1/8
Brocade(config-mif-3/1/5-3/1/8)#link-aggregate configure key 10000
This command sequence changes the key for ports 1/1/1 through 1/1/4 and 3/1/5 through 3/1/8 to 10000. Because all ports in an aggregate link must have the same key, this example forms a multi-slot aggregate link for ports 1/1/1 through 1/1/4 and 3/1/5 through 3/1/8.
Check out the config guide, there's a section on LACP.