Agreed and I would add a comment. A lot of people assume that cut-through
means that the bits have whipped through and disappeared. Actually, Cisco
switches buffer frames regardless of switching mode. This means that the
switch can calculate the CRC, as you say, and keep track of bad CRC counts.
There's another important reason that it must buffer the frame. What if the
output port is a shared (half-duplex) Ethernet? The router must obey
CSMA/CD rules in that case and it may not be able to transmit the bits
right away. The medium might be busy and the frame must be deferred. In
that case, it's a good thing that the frame was buffered.
Priscilla
At 09:27 AM 8/17/01, Peter Van Oene wrote:
>This means that although the switch cannot prevent the forwarding of invalid
>frames since it begins forwarding prior to verifying the checksum, it can
>keep track of the number of errored frames since it does eventually verify
>the checksum. In other words, unlike store and forward switches who verify
>the checksum before forwarding, by the time a cut-through switch realizes
>the frame is bad, the damage is done. This is key for modes like
>adaptive/dynamic cut-through or whatever marketing calls it these days,
>where a switch can use cut-through until a specific error threshhold is
>reached and then dynamically switch to store and forward mode. In practice
>however, the latency variance between modes is so minimal that I believe
>almost all switches use store and forward.
>
>Pete
>
>
>*********** REPLY SEPARATOR ***********
>
>On 8/17/2001 at 6:39 AM Phil Barker wrote:
>
> >Just reading "Layer 2 Switching and Bridging" by Leigh
> >Anne Chisholm.
> >
> >Cut-Through
> >
> >"Since the port does not wait to receive the CRC at
> >the end 'if/of' the frame, it cannot determine the
> >integrity of the data received"
> >
> >Happy with that.
> >
> >"Cut-through switches CAN perform a CRC check as the
> >frame passes through the switch, keeping track of the
> >number of bad frames the port receives".
> >
> >I'm Confused.
> >
> >I'm guessing that some form of CRC checksum can be
> >calculated on the first X-bits of the frame before the
> >cut-through process is allowed rather than the entire
> >frame.
> >
> >Any ideas ?
> >
> >Phil.
> >
> >
> >
> >____________________________________________________________
> >Do You Yahoo!?
> >Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
> >or your free @yahoo.ie address at http://mail.yahoo.ie
________________________
Priscilla Oppenheimer
http://www.priscilla.com
Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=16410&t=16354
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
