[c-nsp] Per Packet vs Per Destination Load Balancing

[Rodney Dunn]

Current packet forwarding implementations don't
do any type of dynamic load balancing (other than
MLPPP).  Packets are forwarded based on some
mechanism (per packet, per destination, per flow, etc.)
with no knowledge of load.  MLPPP is the only one
that takes the load in to account.

So "new traffic does _not_ tend to migrate to the other less used circuit"
is expected.

Rodney


t, Oct 23, 2004 at 04:34:13PM -0700, Christopher McCrory wrote:
> Hello...
> 
> 
> On Fri, 2004-10-22 at 17:21, Chris Strandt wrote:
> > Is there a performance impact or limitation with per destination load 
> > balancing?
> > 
> > It seems to me that with per destination, IOS has to keep track of the 
> > source/destination pair.  Will this slow forwarding, or can this 
> > negativly impact a router forwarding a DDOS attack?
> > 
> 
> (not quite answering you question, but giving you another data point) 
> 
> >From my experience (all of a week:) with balancing two T1 lines (ospf
> +cef).  With per-packet; any traffic can use all the bandwidth.  With
> per-destination; when one link is saturated, new traffic does _not_ tend
> to migrate to the other less used circuit.  On a 2620, I didn't notice
> any significant cpu usage either way.
> 
> 
> > Thanks,
> > -Chris
> 
> -- 
> Christopher McCrory
>  "The guy that keeps the servers running"
>  
> chrismcc at pricegrabber.com
>  http://www.pricegrabber.com
>  
> Let's face it, there's no Hollow Earth, no robots, and
> no 'mute rays.' And even if there were, waxed paper is
> no defense.  I tried it.  Only tinfoil works.
> 
> _______________________________________________
> 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/