misunderstanding of 'backpressure' algorithm

I believe the attached technical report should explain how, and why the part behaves as it does/

-- Rich Seifert Networks and Communications Consulting 21885 Bear Creek Way (408) 395-5700 Los Gatos, CA 95033 (408) 228-0803 FAX

Send replies to: usenet at richseifert dot com

begin 644 TechRept15.pdf M)5!$1BTQ+C$@#27BX\\_3(`TQ,"`P(&]B:@T\\/`TO3&5N9W1H(#$Q(#`@4@TO M1FEL=&5R("],6E=$96-O9&4@#3X^#7-T-PSC*.8Z# M3(`01^-PZ#D-Z?#2-T`J2I:FP*U26M]#$%P,&8;AO'!Q!\\$`0H=GX"&5P,H:P0!7;U1VC\\`3T!,CB"`(]%8XNZ"&$ M$$!E2I`PID>AYR-0I%0;3X;#+QGB/0@N$=F'('1/Y_4`K& M41X1?'_P_:_,$IDEJ[!U#.YX.CNF$@V5`0X@(`UE;F1S=')E86T-96YD;V)J M#3$Q(#`@;V)J#3$T.#,-96YD;V)J#30@,"!O8FH-/#P-+U1Y[J M@YT/]F0+/*`[JT-:NUB_K0C"XK$4"J'I2P6L,SW6.=H\\$F2JAO2TN%![;U MFMR4.S@&ZTRH!U90N4K#(F*I"-^VLK1B`@N?9&[@.H;4A0?5='D]17I/F/D# M'&.8&5,YP(*)T M5HO1FR9J+*ET1V^PZE7$,H_0E8I[E@GO&NAC2YS"NPWJ]5^D8[#JDN%A#&&L M,I7)\\G$9\\E=#,HC1G.5ZB(-S:S4!N-X:9JA4RJHZ0B'(-M"%C-..V9I*=&P8 M`T48"@YZ67LU"L@98Q`4W+,!M0N0R)53/H94U7%5"/8$'M8A74T)T0PAJ:_0 M2IBNX2W$1/:LLYU#)QK,=>0&T\\J/FH5%&P/*N&?(V#F9:NRQ*(,A\\L%8IP2GCP&&/0(#EEY:6]V+J"PWSN="&21

Reply to
Rich Seifert
Loading thread data ...

Hello, All!

I'm reading datasheet for KS8995MA 5-port switch (former Kendin, now Micrel) and got stuck with understanding. Here is abstract from datasheet:

" If back pressure is required, the KS8995MA will send preambles to defer the other stations' transmission (carrier sense deference). To avoid jabber and excessive deference defined in 802.3 standard, after a certain time it will discontinue the carrier sense but it will raise the carrier sense quickly. This short silent time (no carrier sense) is to prevent other stations from sending out packets and keeps other stations in carrier sense deferred state. If the port has packets to send during a back pressure situation, the carrier sense type back pressure will be interrupted and those packets will be transmitted instead. If there are no more packets to send, carrier sense type back pressure will be active again until switch resources free up. If a collision occurs, the binary exponential back-off algorithm is skipped and carrier sense is generated immediately, reducing the chance of further colliding and maintaining carrier sense to prevent reception of packets. "

What I don't take is why to send preambles first, in my opinion it's enought to stop career-sense for awhile to avoid other station's from sending new frames and it also can cease present transmissions?

And for the second, as I know 'back-pressure' isn't standardized, but is there anywhere more-or-less comprehensive description of mechanism?

Thank you very much in advance!

With best regards, Roman Mashak. E-mail: snipped-for-privacy@tusur.ru

Reply to
Roman Mashak

My proposal in the paper calls for discontinuing the preamble transmission *briefly*, to avoid invoking either a jabber or an excessive deferral condition. However, if the carrier is then reasserted quickly (before the 2/3 IFG time), it can still prevent other stations from initiating their transmissions.

-- Rich Seifert Networks and Communications Consulting 21885 Bear Creek Way (408) 395-5700 Los Gatos, CA 95033 (408) 228-0803 FAX

Send replies to: usenet at richseifert dot com

Reply to
Rich Seifert

Hello, Rich! You wrote on Mon, 09 Jan 2006 08:37:07 -0800:

??>> deference). To avoid jabber and excessive deference defined in 802.3 ??>> standard, after a certain time it will discontinue the carrier sense ??>> but it will raise the carrier sense quickly. This short silent time ??>> (no carrier sense) is to prevent other stations from sending out ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ??>> packets and keeps other stations in carrier sense deferred state. If

RS> I believe the attached technical report should explain how, and why the RS> part behaves as it does/ Thank you for document, it's indeed clarified part of doubts. But still I don't understand 'underlined', as I know absence of carrier indicates the media is free and is allowed to transmit data?

With best regards, Roman Mashak. E-mail: snipped-for-privacy@tusur.ru

Reply to
Roman Mashak

No *transmission on the medium*, no carrier sense. Preamble is just a convenient, benign signal to transmit on the medium.

-- Rich Seifert Networks and Communications Consulting 21885 Bear Creek Way (408) 395-5700 Los Gatos, CA 95033 (408) 228-0803 FAX

Send replies to: usenet at richseifert dot com

Reply to
Rich Seifert

Hello, Rich! You wrote on Tue, 10 Jan 2006 08:13:00 -0800:

RS> My proposal in the paper calls for discontinuing the preamble RS> transmission *briefly*, to avoid invoking either a jabber or an RS> excessive deferral condition. However, if the carrier is then RS> reasserted quickly (before the 2/3 IFG time), it can still prevent RS> other stations from initiating their transmissions. Did I understand right: no preamble - no carrier sense ?

With best regards, Roman Mashak. E-mail: snipped-for-privacy@tusur.ru

Reply to
Roman Mashak

(snip)

As far as I know, it was to give the PLL something to sync. to before the actual data arrived.

-- glen

Reply to
glen herrmannsfeldt

Hello, Rich! You wrote on Wed, 11 Jan 2006 08:37:45 -0800:

RS>>> reasserted quickly (before the 2/3 IFG time), it can still prevent RS>>> other stations from initiating their transmissions. ??>> Did I understand right: no preamble - no carrier sense ? ??>>

RS> No *transmission on the medium*, no carrier sense. Preamble is just a RS> convenient, benign signal to transmit on the medium. So, issuing preamble is a way to announce about upcoming transmission ?

With best regards, Roman Mashak. E-mail: snipped-for-privacy@tusur.ru

Reply to
Roman Mashak

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.