Cell Relay Archive[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index][Thread Index][Author Index][Subject Index] Re: Measuring Cell Transfer Delay
In <32826190.102460@news.demon.co.uk>, andy@gargoyle.demon.co.uk (Andrew Valentine) writes: >Question >---------- > >When SVCs are used it is possible to request in the SETUP that the connection >has a max CTD of X. As the SETUP traverses the network towards the >destination each switch adds its CTD to the accumalated value for the >connection so far. The question I have is how do switch manufacturers >measure the CTD (that is used during SVC setup) for their switch. Is it >first bit in last bit out? > Not sure. I would guess it is First bit in/ First bit out usually. >If I wanted to measure the actual CTD performance of a switch would I measure >last bit in first bit out so that I do not get input serialization times? This is not a good measure as it could lead to NEGATIVE values if a switch was a cut-through switch & very efficient at cell processing. Likewise First bit in last bit out will overstate the delay because by the time the switch has transmitted the last bit the first bit will be received by the next switch in line. (actually almost all the bits will be received). > >Many thanks > >Andrew Valentine > > > Francis Turner Madge US - All opinions are my very own though fturner@madge.com |
|