Re: OT crossover cable speed
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Switches don't really have collisions per-say - in an ethenet sense. Most
consumer small swithes experience what's called head-of-line blocking,
however. This is the case when a frame is queued to go out a particular
port which is busy, behind this frame (in the queue) is another packet
waiting to go out another port which happens to be free this cycle. It
wont be processed, unfortunatly, until the packet ahead of it can go out.
Larger switches Cisco 6500, Foundry, Extreme, ...etc don't have this
problem by having seperate queues for evey port.
On Wed, 27 Mar 2002, Nathan E Norman wrote:
> On Tue, Mar 26, 2002 at 11:09:00PM -0500, Jason Healy wrote:
> > At 1017245979s since epoch (03/26/02 22:19:39 -0500 UTC), Crispin Wellington wrote:
> > > Of course theres collisions. But switch <-> Computer has collisions
> > > too.
> >
> > Unless you're running in full duplex (both crossover and switches
> > support this mode, so long as the NIC/switch at each end both support
> > it and are configured correctly). Then there's no collisions, and
> > you have a nice fast connection.
>
> Well, full duplex prevents collisions from host <-> switch. But
> switches do have collisions internally when two ports want to talk to
> a third port at the same time. It's difficult to know whether this is
> happening unless you have a managed switch.
>
> --
> Nathan Norman - Micromuse Ltd. mailto:nnorman@micromuse.com
> Gil-galad was an Elven-king. | The Fellowship
> Of him the harpers sadly sing: | of
> the last whose realm was fair and free | the Ring
> between the Mountains and the Sea. | J.R.R. Tolkien
>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: Made with pgp4pine 1.76
iD8DBQE8ogpAWMfBXOjbFQcRAsVKAJ0UsMl1g11Bl4EZJnxMZCgV8wO2hQCcCtdR
HIJASlZjvjn16eryUNea1W0=
=XN+9
-----END PGP SIGNATURE-----
--
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: