RSPAN limits on Native IOS (was Re: [nsp] number of netflow entries on cat6K w/sup2)

Michael Sinatra michael at rancid.berkeley.edu
Wed Sep 24 10:29:04 EDT 2003



On Wed, 24 Sep 2003, ken lindahl wrote:

> At 12:26 AM 9/24/2003, sthaug at nethelp.no wrote:
> >You are hopefully aware of the fact that 6K native supports RSPAN in
> >reasonably new IOS versions?
>
> hmm, i looked into this about 6 weeks ago and did not see what i
> needed. iirc, there was a limit to the number of ingress and
> egress source ports (the application needs to see traffic in both
> directions). sounds like i should look again.

I don't believe anything has changed with respect to RSPAN limits during
that time.  (There doesn't appear to have been a significant new Native
IOS release, either.)  The problem is, within the RSPAN session limits
(which appear to be roughly the same on hybrid vs. native), native imposes
a greater limit on the number of TX ports that can be included in a single
session that's much more stringent than what hybrid (catos) imposes.  (We
have done testing--catos has been working fine for some time, but native
IOS won't let us configure more than one TX port in a session.)  Because
of our current needs wrt RSPAN, this limitation is significant--we cannot
easily jury-rig the topology so that all of the relevant traffic can be
captured using SPAN or RSPAN on RX-only monitored ports.  It requires a
more substantial redesign, which will be really easy because we have tons
of free time, just like everyone else.

It sux because I'd really like to go native on all of our boxes, but the
two that have the special RSPAN needs are stuck at hybrid.

michael



More information about the cisco-nsp mailing list