[c-nsp] cef load-sharing per-packet bizzare behavior

[Yasser Aly]

Rodney,

  Interesting feedback you provide. I have a situation that appears to be
similar to me. Having a FastEthernet interface on 7507 where DCEF is
applied. The Fast Ethernet has a sub-interface running dot1Q. Doing show
interface f1/1/0 shows traffic only in the input rate counter, output rate
counter is always Zero, this is not true as monitoring this interface via
MRTG shows traffic flow in both directions.

Any thoughts ?

Regards,
Yasser 

-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net
[mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Rodney Dunn
Sent: Wednesday, September 15, 2004 11:40 PM
To: Pete Templin
Cc: cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] cef load-sharing per-packet bizzare behavior

Yes....fifo is always distributed once dCEF is enabled.
That's why I say make them all FIFO unless you know you need some fancy
queueing of some sort and that way you don't have to worry about it.

99% of the time the customers don't even notice a difference with WFQ off
vs. fifo but I have seen a couple instances where it is noticeable with low
speed links. 

Anything above 2Mbps is FIFO by default anyway so you are good to go on
those.  It's the 2 Mbps or less that bites.

Rodney

On Wed, Sep 15, 2004 at 03:28:42PM -0500, Pete Templin wrote:
> Rodney Dunn wrote:
> 
> > Oh..and one thing I did just remember during lab testing.
> > 
> > Make sure if these are T1's you either do:
> > fair-queue
> > or
> > no fair-queue
> > 
> > so it's no doing RSP based WFQ and the packets are dCEF switched.
> 
> Any tricks for other media?
> 
> What should Ethernet interfaces show as their queueing strategy (i.e. 
> is "fifo" distributed)?
> 
> Pete
_______________________________________________
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/