From: Joel W. Ekis (jekis@xxxxxxxxx)
Date: Mon Mar 13 2000 - 15:51:39 GMT-3
However, the incorrect assumption that immediately follows is: 'If a switch is
non-blocking, then all packets will get through the switch.' I hear this entir
ely too often.
This is rarely the case. Most network designs have many ports talking to 1 or
a few uplink ports. If you have more than 10 FE ports talking wire-rate to a s
ingle GE port, I don't care if your internal switch fabric is 10x beyond non-bl
ocking, you will lose packets. That is the whole reason why buffers and QoS ar
e necessary.
Sorry, I'll get off my soapbox now.
Joel
At 02:43 PM 3/13/2000 +0000, Lee, Vaughan wrote:
>As lifted off the front page of this week's Network News (UK paper):
>
>'A switch has a non-blocking architecture if its internal backplane speed is
>equal to, or greater than, the sum of the speed of all the ports connecting
>to the switch. For example: a non-blocking switch with 24 100Mbps ports
>will have a backplane speed of 2.4Gbps. This means that network traffic
>shouldn't reach a bottleneck when it hits the switch.'
>
>Short, sharp, to the point.
>
>Regards,
>Vaughan
>
>Vaughan Lee, Network Specialist.
>CCIE #5250
>debis IT Services (UK) Ltd.
>E-mail: vaughan.lee@debis.co.uk
>Direct Phone: 01908 279561
>Direct Fax: 01908 279061
>http://www.debis.co.uk/
>
>
>
>-----Original Message-----
>From: Eddie Parra [mailto:eparra@rexallsundown.com]
>Sent: 13 March 2000 13:55
>To: CCIE Group Study
>Subject: Blocking and non-blocking switches???
>
>
>Can someone tell me what the difference is between a "blocking"
>and "non-blocking' switch?
>
>-Eddie
This archive was generated by hypermail 2.1.4 : Thu Jun 13 2002 - 08:23:05 GMT-3