ICMP ping effecting network flow?

Everyone,

I was told by someone at a systems conference about 10 years ago that if
you ping a server over the Internet while downloading a file from it,
your download will take place quicker. I thought it was bogus then and
still kind of lean towards that. I see no difference from some simple
tests I've tried. But, perhaps there's something to do with a common
design of routers and traffic flows that I just don't get. The guy who
told me got a knowing nod from someone else and neither were in a
position to just be pulling legs. Were they both wrong and just trying
to sound smart? Any (real) info would be much appreciated. It's
tickled the back of my mind for a long time.


Roy

Roy L Butler <(E-Mail Removed)> wrote:
> I was told by someone at a systems conference about 10 years ago
> that if you ping a server over the Internet while downloading a file
> from it, your download will take place quicker. I thought it was
> bogus then and still kind of lean towards that. I see no difference
> from some simple tests I've tried. But, perhaps there's something
> to do with a common design of routers and traffic flows that I just
> don't get. The guy who told me got a knowing nod from someone else
> and neither were in a position to just be pulling legs. Were they
> both wrong and just trying to sound smart? Any (real) info would be
> much appreciated. It's tickled the back of my mind for a long
> time.

I'd chalk it up to leg-pulling. The only way I could see a ping
helping a transfer would be if it was _just_ enough to help keep the
NICs' interrupt throttles from being too aggressive.

rick jones
--
The glass is neither half-empty nor half-full. The glass has a leak.
The real question is "Can it be patched?"
these opinions are mine, all mine; HP might not want them anyway...
feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...

On Thu, 11 May 2006, in the Usenet newsgroup comp.os.linux.networking, in
article <e40gdr$5l1$(E-Mail Removed)>, Roy L Butler wrote:

>I was told by someone at a systems conference about 10 years ago that if
>you ping a server over the Internet while downloading a file from it,
>your download will take place quicker. I thought it was bogus then and
>still kind of lean towards that. I see no difference from some simple
>tests I've tried.

Anecdotal evidence - no observed difference. Why? Because the services
have nothing to do with each other. On the server, the 'ftp' or 'web'
server is an application running at the top of the networking stack. With
most operating systems, 'ping' (ICMP Echo) is a network service that is
part of the networking stack. At the client side, 'ping' is normally
just another application - which if anything is stealing CPU cycles that
could ordinarily be used for something more useful.

>But, perhaps there's something to do with a common design of routers and
>traffic flows that I just don't get.

As far as the routers are concerned, it's just another packet. An ICMP
packet _MAY_ be processed at a lower priority than TCP or UDP, because
ICMP is more used as an error, or diagnostic, while the others are
carrying "useful" data.

>The guy who told me got a knowing nod from someone else and neither were
>in a position to just be pulling legs.

Why not? Are you the "boss" that they dare not lie to? Or was their boss
present and would not take kindly to such humor?

>Were they both wrong and just trying to sound smart? Any (real) info
>would be much appreciated. It's tickled the back of my mind for a long
>time.

Ten years ago - to late to do anything. The right thing to do AT THE TIME
would have been to ask them to explain why they felt this to be true.

Old guy