The MPLS WG Archive

[Curtis Villamizar <curtis@workhorse.fictitious.org>]

In message <4B6D09F3B826D411A67300D0B706EFDE84A482@nt-exch-yow.pmc-sierra.bc.ca
>, Shahram Davari writes:
> Curtis,
> 
> > John,
> > 
> > This makes no sense at all.  In the ICMP part of LSP Ping, there is no
> > load on the egress CPU because the forwarding card ASICs handle ICMP.
> > There is a one for one relationship between the number of packets any
> > ingress sends and the number it receives (minus any loss) so if an LSR
> > CPU is loaded too heavily, it can reduce its sending rate (and
> > expected reception rate) and reduce its own CPU load.
> > 
> > If an egress is overloaded with OAM, there may be *NO* errant LSP,
> > just too many perfectly fine LSPs each sending just a little too much
> > probe traffic.  There are limits to what you can get a microprocessor
> > to do in a given amount of time.
> > 
> > Curtis
> >
> 
> You are comparing apples to oranges. In this debate you assume that the ICMP 
> pings
> are processed in hardware but the CV packets are processed in software ! ther
> efore
> you conclude that the egress can't be overloaded with ICMP but it can with CV
> . Strange
> logic!
> 
> -Shahram


Its apples to apples.  I'm assuming the ingress which sends the ICMP
processing on reception does the same amount of processing that is
required to process an MPLS-OAM packet.

The difference is the ICMP receiver is the sender of the same ICMP and
can easily reduce its load if it needs to do so.

Taking an ICMP echo request and making an ICMP echo reply is
mechanical and can be done at or near full forwarding rate or at least
at a very high rate with no impact on forwarding.  This is not a
speculation about performance of a proposed protcol feature.  All
successful routers have done that for years.

Curtis