[c-nsp] mpls hw limits on 7206

Rodney Dunn rodunn at cisco.com
Thu Jun 29 13:09:28 EDT 2006


I'm on PTO starting tomorrow for a week.

I don't know the answer or I would have responded
already. I'll try and dig it up when I get back.

Rodney

On Thu, Jun 29, 2006 at 07:00:47PM +0200, Gert Doering wrote:
> Hi,
> 
> On Thu, Jun 29, 2006 at 02:36:42PM +0100, David Freedman wrote:
> > Yes, if you are implementing ethernet L2VPNs and wish to allow your 
> > customers to have the ability to carry standard "ip mtu" (1500B) inside 
> > their ethernet frames, then steer clear of the following interface types 
> > for the uplink to the core / other MPLS routers:
> > 
> > - PA-FE-TX
> > - PA-2FE-TX
> > - PA-FE-FX
> > - IO/FE
> > - IO/2-FE
> 
> Have you found evidence in the meantime that the 2FE boards (intel-based)
> have the very same 1530-byte-problem as the FE boards (digital based)?
> 
> Rodney, can you comment on that?  As it seems to be a chipset issue, 
> using a different chipset might change things...

See above.

> 
> BTW: is there a way to actually notice if this problem strikes?  I'm
> asking because I recently set up an EoMPLS link that crosses a IO-2FE
> board, configured to "mpls mtu 1530" on 12.2(18)S12 - and transporting
> full-sized ethernet frames plus two labels (1518+4+4+18 = 1544 on-the-wire)
> doesn't seem to cause any issues.  I regularily see about 90 Mbit/s of
> TCP traffic go through, and if there is packet corruption, the TCP 
> endpoints would certainly notice.

No. Usually just pings will fail. The "data corruption" part is junk
I think. It's more about what they have tested and certify. They had
fould some issues where the chip could cause corruption for larger size
frames but that wasn't ever chased down probably.

> 
> 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
> _______________________________________________
> 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