etherchannel question PoXXA

[simionov.adrian]

I configured a etherchannel with 4 port (this will be interface portchannel 20), all is ok.

Last command used is, on the physical ports, no shutdown.

When this command is issued, the switch creates another virtual port-channel interface, the name is 20A, and one physical port is here, and 3 physical port are on the correct port channel interface 20.

I do not have access to the config on the other side of the configuration. Switch is VSS 6500 box.

Any idea where to search more info regarding this? Is the first time when I see this behavior.

[Richard Michael]

Hello,

if the ports were not added properly to the portchannel then you may see a subinterface 20A to be created (Wierd logic )

=> Check the configs of the ports and the portchannel at both the ends It should be same

=> Remove the port channel and put the interface to default

=> Create the port channel again and try to bring up the ports

If its still creating 20A or 20B and If you see any error messages such as some attributes or aggregators incompatibility, Please post the error message with the sh ver,sh mod and the configs of the portchannel and the physical ports.

Thanks,

Ricky Micky

Pls rate if the content was useful

[simionov.adrian]

The answer is:

http://www.cisco.com/en/US/tech/tk389/tk213/technologies_configuration_example09186a0080094470.shtml#po1a

[Richard Michael]

Hi,

Happy that you figured it out

Secondary channels are usually created due to a config mismatch or an incompatible feature existed between the ports. The reason for the incompatibility would usually be given as a syslog message. Note that if you did not see the message on the "near side" chassis make sure you also look at the far-end chassis that the partner port is located. The incompatibility may be there instead. Clear the incompatibility and all ports should bundle within the same port channel. Incompatible ports might be not the exact interface type or any other issues that may exist on these ports or ports on the other switch.

Now comes the wierd stuff

The reason we ended with Po2A is that when the etherchannel was brought up, one link was access the other was trunk, this is the kind of inconsistency that brings up secondary aggregators which are not configurable.

It comes down to a race scenario between LACP and Dynamic Trunking. I not 100% sure which protocol goes first one link of the channel goes trunking faster than the other link from the channel.  Then LACP detect the missmatch and choose that a secondary aggregator should be build and when the valid condition come back the single aggregator is not restored.

Thanks,

Ricky Micky

*Rate if this helps