[c-nsp] CBAC unsupported on the 7500's

Rodney Dunn rodunn at cisco.com
Mon Oct 15 10:09:20 EDT 2007


I hear what you are getting at and I agree for the most part.
I push for the "right thing" as much as possible.

But just for the record...

On Mon, Oct 15, 2007 at 09:50:28PM +0800, Adrian Chadd wrote:
> On Mon, Oct 15, 2007, Joe Maimon wrote:
> > So after over a year of issues with ACL's/CBAC I get the word.
> > 
> > "CBAC is unsupported on 7500 series"
> > 
> > However, a google search of "CBAC" "7500" turns up:
> > 
> > http://www.cisco.com/en/US/products/hw/routers/ps359/prod_brochure09186a00800886e4.html
> > 
> > This isnt very funny. Is it only the 7500 platform which gets 
> > retroactively unsupported features? Why isnt it in release notes? 
> > Howcome there are no CLI prompts for unsupported features?
> 
> I don't think its just the 7500. There's plenty of examples of supported
> features becoming unsupported because of "bugs" in the hardware.

That's not the case here. It's a matter of testing and support. This box has
been out there with those feature sets for years. And this just now comes up
tells me not many people wanted/needed/used it on the 75xx. Given that box is
in sustaining mode don't expect anything to complex to be done with it.
The biggest problem here (and I agree with Joe) is we didn't block it in the CLI
and it's even referenced in some marketing document.

> 
> Eg: The Cisco 3550 12.1 vs 12.2 MQC policy map matching rules; I think
> the 3550's even did netflow at some point(!); the Cisco PXF based stuff
> becoming "unsupported" because (and I'm semi-quoting from memory here)
> they couldn't resolve all the bugs?

No that's not true. It's not that they "couldn't" resolve all the bugs it
was that customers were asking for more features at speeds that were reachable
with new CPU architectures so with the ability to disable PXF on that box they
decided not to fix any more PXF bugs on those couple of platforms (it was only
the 7401 and NSE100). PXF is still being used on other platforms and bugs are being
fixed. I'm not defending the shorter lifecycle decision but I was part of those
discussions and from the overall view they made the right decision from all sides.

Others will disagree I'm sure.


> 
> Who bought NPE-G1's and got pissed off that you couldn't really use
> the second CPU even though Cisco kept saying it was groundbreaking? :)

They tried it for a certain subset of feature acceleration and it worked
pretty well. But again the feature complexity became an issue because they
had no way to block all the features that were not supported on the second CPU
so it caused more problems than it helped. They (the BU) even tried to put in
a strict enforcement policy around deployments and that didn't even work.

Rodney


> 
> 
> 
> Adrian
> 
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/


More information about the cisco-nsp mailing list