Trunking Configuration Problem - ErrDisable

[cindylee27]

Hi Experts,

I encountered one problem as below..

ERR_DISABLE: channel-misconfig error detected on Po1, putting Fa0/23 in err-disable state

When I tried to configure etherchannel group for 2 fastethernet port of fa0/23 and fa0/24.

Fa0/23 is connected to Core Switch 1.

Fa0/24 is connected to Core Switch 2.

Both are one channel-group 1 mode on.

Config as below:

On Core Switch 1:

Int fa2/5

Description To Dist Sw__1_Fa0/23

Switchport trunk encapsulation dot1q

switchport mode trunk

speed 100

duplex full

On Core Switch 2:

Description To Dist Sw_1_Fa0/24

Int fa2/5

Switchport trunk encapsulation dot1q

switchport mode trunk

speed 100

duples full

On Dist SW_1,

Interface Port-Channel 1

Switchport trunk encapsulation dot1q

Switchport mode trunk

...

...

...

Int fa0/23

switchport trunk encapsulation dot1q

switchport mode trunk

speed 100

duplex full

channel-group 1 mode on

...

...

...

Int fa0/24

switchport trunk encapsulation dot1q

switchport mode trunk

speed 100

duplex full

channel-group 1 mode on

...

...

...

Note: When I leave one cross cable not connected, it doesn';t give any problem.

Only when I connect both cross cable to core Switch 1 and 2, then it will go to Err-Disable mode for the port fa0/23 and fa0/24.

Thanks in advanced.

[Edison Ortiz]

To avoid this err-disable message, you need to shutdown both switchports at each link-partner before entering channel-group x mode x

Once you are done with the channel-group configuration, you can issue a 'no shut' under the interfaces.

[Kevin Dorrell]

Etherchannel bundles must always go between the same two switches. You cannot have one link of a bundle going to one switch, and another to a different switch.

Furthermore, the two links would need to be bundled at the core switch end as well.

Kevin Dorrell

Luxembourg

[cindylee27]

Guys,

Thanks for your reply.

Edison, Did u mean that I need to do a shut on interface fa0/23 and fa0/24,before applying the command : "channel-group 1 mode on" to prevent this problem?

Can I say that once applied the command above, I can do a "no shut" and will solve the problem?

Kevin, If that is the case, How can I configure 2 redundancy links to both Core Swithes,in case Core Switch 1 or 2 fails.. ?

Thanks again..

[Edison Ortiz]

I missed the part of the destination device being different.

As Kevin pointed out, you can't have 2 sources switchports going to 2 different destination devices in an etherchannel.

[cindylee27]

So,how can I create 2 connections to 2 diff Core Sw which both are trunked to each other ? CS1 -> CS2 (etherchannel trunk ).

Will I cause a looping and how to configure redundancy for this type of connection/??

Thanks.

[Edison Ortiz]

Trunk is allowed on both connections while etherchanneling is not, there are 2 different features.

Etherchannel provides bandwidth aggregation while trunking provides the ability to send multiple Vlans over a single link.

You don't need to etherchannel in order to get redundancy. Just configure the trunking and plug the links from the leaf switch up to the Cores.

[cindylee27]

Thanks Edison..

I read somewhere there looping problem might happens if this is the case..Can I say that the looping is being taken care off by STP automatically or is there any configurations that you can point me to ?

Is that means tat I can configure without port-channel 1 for the fastethernet 0/23 and 0/24 ?

and connect both to Core Switch 1 and 2 respectively?

Thanks again.

[Edison Ortiz]

The looping is prevented with STP. Depending on how your STP is configured, one of those ports will be blocked.

For instance, if Core 1 is the STP Root for all Vlans, and port 0/23 is connected to this Core - will be set to FWD mode while 0/24 will be set to BLK mode (you can verify this with show spanning-tree command on the leaf switch).

If you have multiple Vlans, you can load-balance them by assigning Core1 STP Root half of the Vlans while Core2 STP Root for the other half.