[c-nsp] ip route-cache

[Rodney Dunn]

<snip>

> (Unfortunately, we're unlikely to move to 12.2(25)S any time soon.  The
> images have grown quite a lot, and 256 Mb boxes with BGP and 12.2(18)S
> are already down to 40 Mb "Free" [vs. 60 Mb on 12.2(14)S!] - so I fear
> 12.2(25)S is going to bring with it an upgrade to 512 Mb DRAM, which
> needs budget...)

I understand that and while we would like to be able and provide such
support enhancements in old code sometimes they require infrastructure
work to get them.  That's why they can't be in older/smaller images.

The goal is to think outside the lines about what *would* be
nice from a support perspective and try to build things to
help the customer.  So if you have ideas, please do take a few
minutes and send me an email.

On Thu, Oct 14, 2004 at 07:43:32PM +0200, Gert Doering wrote:
> hi,
> 
> On Thu, Oct 14, 2004 at 01:17:45PM -0400, Rodney Dunn wrote:
> > Post 12.2(25)S code is totally different from a switching
> > path perspective.
> [..]
> > You can do it per interface and match on an ACL with
> > very little to no performance impact.  The other interfaces
> > that don't have the debug applied take *no* performance
> > hit with the debug enabled.  We view this as a small step
> > forward in helping the user debug packet forwarding problems.
> 
> COOOOOOOOOOOOOOL.  Sorry for yelling, but that's GREAT stuff.
> 
> This is exactly what "debug ip packet" never properly did - there's
> a "debug ip filter" but that never did what I tried to achieve, and also
> it only worked with process switching.
> 
> (Unfortunately, we're unlikely to move to 12.2(25)S any time soon.  The
> images have grown quite a lot, and 256 Mb boxes with BGP and 12.2(18)S
> are already down to 40 Mb "Free" [vs. 60 Mb on 12.2(14)S!] - so I fear
> 12.2(25)S is going to bring with it an upgrade to 512 Mb DRAM, which 
> needs budget...)
> 
> [..]
> > If you have additional ideas on things that you think
> > would make the box more troubleshootable in production
> > feel free to forward the ideas to me.
> 
> I'll certainly give this a good thought :-)
> 
> A fair number of things have already appeared (like "ACL logging telling
> you the sequence number that matched", so you know *which* statement
> dropped the packet, without going through all of them).
> 
> gert
> -- 
> USENET is *not* the non-clickable part of WWW!
>                                                            //www.muc.de/~gert/
> Gert Doering - Munich, Germany                             gert at greenie.muc.de
> fax: +49-89-35655025                        gert at net.informatik.tu-muenchen.de