Network Link Aggregation (i.e. Etherchannel)

[NETWizz]

Link Aggregation is great for redundancy. Basically, it improves bandwidth and makes a loopless topology whereby spanning-tree will not shutdown the ports.

What you need to know is that there are two (2) protocols predominantly used. Cisco uses their proprietary protocol, PAgP, and every other vendor I have ever seen uses LACP.

Here is how to setup a Link Aggregation on a couple Cisco Switches:

On multiple interfaces, add them to a channel group.:
interface FastEthernet0/1
channel-group 1 mode desirable
!
interface FastEthernet0/2
channel-group 1 mode desirable
!

Usually, it is easier to do this as a range:
Switch(config)#int range fa0/1 - fastEthernet 0/2
Switch(config-if-range)#channel-group 1 mode desirable


This creates a virtual interface:
interface Port-channel 1

!

Here is how to get a summary:

Switch#sh etherchannel summary
Flags: D - down P - in port-channel

I - stand-alone s - suspended
H - Hot-standby (LACP only)
R - Layer3 S - Layer2
U - in use f - failed to allocate aggregator
u - unsuitable for bundling
w - waiting to be aggregated
d - default port


Number of channel-groups in use: 1
Number of aggregators: 1



Group Port-channel Protocol Ports
------+-------------+-----------+----------------------------------------------


1 Po1(SU) PAgP Fa0/1(P) Fa0/2(P)




What you need to know is that to communicate with other Vendors, just use the keyword, "Active" instead of desirable because they use LACP.

Now the Terminology is different. many vendors refer to Link Aggregation as LAGs. Cisco likes to use the terminology "etherchannel," but it is the same thing.

Here is an example of how to do it on a Brocade ICX Family, which essentially runs the FastIron style IOS from the Foundry Days (with newer 8.x code):

!
!
lag technibble static id 1
ports ethernet 1/1/1 to 1/1/2
primary-port 1/1/1
deploy
!


That's it... The first two ports now participate in a LAG. They use LACP and will work perfectly fine even with Cisco provided you set the Cisco device to use LACP instead of its proprietary PAgP.


ICX6430-48P Switch#sh lag

Total number of LAGs: 1
Total number of deployed LAGs: 1
Total number of trunks created:1 (28 available)
LACP System Priority / ID: 1 / cc4e.24cd.2580
LACP Long timeout: 90, default: 90
LACP Short timeout: 3, default: 3

=== LAG "technibble" ID 1 (static Deployed) ===
LAG Configuration:
Ports: e 1/1/1 to 1/1/2
Port Count: 2
Primary Port: 1/1/1
Trunk Type: hash-based
Deployment: HW Trunk ID 1

Port Link State Dupl Speed Trunk Tag Pvid Pri MAC Name
1/1/1 Up Forward Full 1G 1 No 1 0 cc4e.24cd.2580
1/1/2 Up Forward Full 1G 1 No 1 0 cc4e.24cd.2580


Now, I usually use them in a stacking configuration such that I an loose an entire switch without loosing the configuration or the entire LAG.

It is exactly the same thing except the stacked switches enumerate their ports like this 1/x/x first switch... 2/x/x second switch etc.

 

[trevm999]

Now, I usually use them in a stacking configuration such that I an loose an entire switch without loosing the configuration or the entire LAG.

It is exactly the same thing except the stacked switches enumerate their ports like this 1/x/x first switch... 2/x/x second switch etc.

I like the idea of that.

 

[trevm999]

How would this affect port mirroring?

 

[NETWizz]

How would this affect port mirroring?

Not sure how it would affect that. Without doubt, you would need enough bandwidth the mirror the lag, but it should work.

 

[JaTom]

Thank you for posting this! I have been looking into this technology recently and seeing this post has got me going back to those rabbit holes! I think this will help with my NAS solutions.