I am not quite sure how an ATM switch becomes a "packet shredder," or even what a "packet shredder" is...
An ATM switch shreds a 1500 byte IP packet into 29 ATM cells of 53 bytes each.
At the risk of being excessively pedantic, the host ATM interface, not the switch, segments the SDU into cells. My point was that the term "packet shredder" doesn't seem to have any very precise meaning beyond perhaps the writer's desire to use a catchy phrase to indicate an attitude towards ATM. Some people use it to describe the segmentation function of ATM interfaces, other use it to describe switches which don't do packet discard.
If an ATM has a 5% loss, i.e. 1 in every 20 cells is lost, then 0% of the IP traffic will get through. ...
Most anything with 5% loss is broken. -tjs
On Mon, 29 Jul 1996, Tim Salo wrote:
I am not quite sure how an ATM switch becomes a "packet shredder," or even what a "packet shredder" is...
My point was that the term "packet shredder" doesn't seem to have any very precise meaning beyond perhaps the writer's desire to use a catchy phrase to indicate an attitude towards ATM.
I have always interpreted it as meaning a protocol below the IP level which breaks IP packets up into fragments for transport.
If an ATM has a 5% loss, i.e. 1 in every 20 cells is lost, then 0% of the IP traffic will get through. ...
Most anything with 5% loss is broken.
Just a worst case scenario intended to illustrate that in the case of ATM packet shredding, losses at the lower level (ATM) can have an effect at the higher level (IP) that is roughly 20 times as great. By now people have realized this and are attacking the problem many different ways, but the problem still does exist and is an artifact of the relationship between IP packet sizes and ATM cell sizes. Michael Dillon - ISP & Internet Consulting Memra Software Inc. - Fax: +1-604-546-3049 http://www.memra.com - E-mail: michael@memra.com
participants (2)
-
Michael Dillon -
salo＠msc.edu