RE: VSS dual active detection mechanisms

In the deployments I have done:

PaGP+ DA - I prefer this method however I really dislike the idea that a
downstream switch can tell the core what to do. Even though I have not seen
it yet, there always could be a code change or add that causes a switch to
have issues where one of the VSS switches ultimately shut its ports. Not
good.

Fast Hello - Few things with this one. First off, you use this if your
downstream switches do not support PaGP+ - say if you're connecting to UCS
Fabric interconnects or switches from other vendors. Second, you will need
SXI1 or later code. Lastly, you burn a port on each switch.

Hope that helps.

Travis

-----Original Message-----
From: nobody_at_groupstudy.com [mailto:nobody_at_groupstudy.com] On Behalf Of
Joseph L. Brunner
Sent: Thursday, October 21, 2010 7:58 AM
To: Koen Zeilstra; ccielab_at_groupstudy.com
Subject: RE: VSS dual active detection mechanisms

Can you tell me why port channels should be dynamically configured Koen?

If not, you should not be using them.

-----Original Message-----
From: nobody_at_groupstudy.com [mailto:nobody_at_groupstudy.com] On Behalf Of Koen
Zeilstra
Sent: Thursday, October 21, 2010 8:04 AM
To: ccielab_at_groupstudy.com
Subject: OT: VSS dual active detection mechanisms

Hi group,

Here's a poll just out of curiousity.

Which of the three mechanisms Cisco provides is your favorite (or least
annoying :-)) .

1. dual active detection using PAgP
2. dual active detection using IP BFD
3. dual active detection using fast hello's

I personally dislike the fact that 1. is proprietary and as a result not an
option in a multivendor environment. Further more I would prefer
portchannels to be statically configured instead of dynamic.

2 en 3 require a direct layer2 ethernet connection, which is a waste of
ports (especially in a chassis which has only 10 G interfaces).

What's your opinion? Maybe there are more methods that I have overlooked.

thanks!

Cheers,

Koen

Blogs and organic groups at http://www.ccie.net
Received on Thu Oct 21 2010 - 12:07:10 ART