Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: Ethernet Characteristics re: packet loss




My two cents,

The packet drop does not happen inside the processor either even when it is
running at 100% busy.  The flow control is managed by the TCP window. It is
known that using a motherboard with dual Pentium III running at 100%, one
can only get 50% to 70% of the gigabit Ethernet.  Hence, everyone is looking
at a TCP Offload Engine (TOE) which is an adapter offloading the TCP
processing.  TOE is a must for 10 Gb Ethernet.

Y.P. Cheng, CTO, ConnectCom Solutions Corp.

> -----Original Message-----
> From: owner-stds-802-3-hssg@xxxxxxxx
> [mailto:owner-stds-802-3-hssg@xxxxxxxx]On Behalf Of Ed Grivna
> Sent: Monday, March 19, 2001 12:46 PM
> To: StoutT@xxxxxxxxxxxx; stds-802-3-hssg@xxxxxxxx; tripathi@xxxxxxxxxxxx
> Subject: RE: Ethernet Characteristics re: packet loss
>
>
>
>
> This is for a collision environment.  One you go to swicthed, this
> problem goes away, and is limited by the processing capability of the
> source/destination and any swithc elements between.
>
> Regards,
>
> Ed Grivna
> Cypress Semiconductor
>
> >
> > The 40% utilization characteristics reminds me of coax medium. I do not
> > think
> > it is true of full duplex media, unless I am missing the
> context some how.
> >
> > Regards,
> >
> > Devendra Tripathi
> > VidyaWeb, Inc
> > 90 Great Oaks Blvd #206
> > San Jose, Ca 95119
> > Tel: (408)226-6800,
> > Direct: (408)363-2375
> > Fax: (408)226-6862
> >
> > > -----Original Message-----
> > > From: owner-stds-802-3-hssg@xxxxxxxx
> > > [mailto:owner-stds-802-3-hssg@xxxxxxxx]On Behalf Of Stout, Tony
> > > Sent: Monday, March 19, 2001 8:51 AM
> > > To: 'stds-802-3-hssg@xxxxxxxx'
> > > Subject: Ethernet Characteristics re: packet loss
> > > Importance: Low
> > >
> > >
> > > I have a quick question.... someone made a statement that it is a
> > > characteristic that once you go above 40% LAN capability, the you
> > > will begin
> > > to lose packets.  This doesn't seem to hold water with me.  I recently
> > > performed a test whereby we injected packets at 100%, 75%, 50%,
> > > 25% 10% & 5%
> > > network capability.  We had 2 laptops, 1 desktop (PIII dual
> processor), a
> > > LAN Analyzer, and a SmartBits 200 generator -- at 100% both the
> > > laptops 40%
> > > of the packets we were sending -- the desktop captured 70% --- in
> > > both cases
> > > the application running on the platforms was using 100% of
> the processing
> > > time. -- my suspicion is that this is where the packets are being
> > > loss -- in
> > > the processing arena.
> > >
> > > Can you shed any light on this?
> > >
> > > Thank you,
> > >
> > > Tony Stout
> > > JITC
> > > Information Systems Test Director
> > >
> > > Phone:  301-744-2690
> > > Mobile: (301) 399-9078
> > > Fax:  301-744-2688
> > >
> > > JITC Webpage: http://misd.ncr.disa.mil/misd
> > >
> > >
> > >  <<Tony Stout (E-mail).vcf>>
> > >
> >