I don't catch what really mean of goodput? What is different from throughput? Let me know any goodmark about it if you have gotten! -- Best regards, James K. (txdiversity@hotmail.com, http://home.naver.com/txdiversity) - Any remarks, proposal and/or indicator to text would be greatly respected. - Private opinions: These are not the opinions from my affiliation.
Goodput: show me a goodmark about it!
Started by ●November 29, 2003
Reply by ●November 29, 20032003-11-29
James K. wrote:> I don't catch what really mean of goodput? > What is different from throughput? > Let me know any goodmark about it if you have gotten!"Goodput" is not a word I know. Can you supply a context? that would help. Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●November 29, 20032003-11-29
"James K." <txdiversity@hotmail.com> wrote in message news:<bqacjb$42f$1@news1.kornet.net>...> I don't catch what really mean of goodput? > What is different from throughput? > Let me know any goodmark about it if you have gotten!Goodput is defined as the rate of transmission of useful data. It is usually expressed in bytes per second. Throughput is the rate of transmission of data (useful data + header information). For example IPv4 datagram packet has 20 bytes of header information. Also assume the datagram has 10 bytes of data. If it is transmitted in a second, then goodput is 10bytes/sec, throughput is 20+10 = 30bytes/sec. Some people consider throughput and goodput as the same (rate of transmission of useful data). Guess it depends on how you define them. -Jay
Reply by ●November 30, 20032003-11-30
Goodput is used to describe the actual throughput of TCP packets, taking
into
account the behavior of TCP backoff, slow start, congestion, and duplicate
packets.
If a packet is lost, TCP interprets the loss as congestion, and throttles
back
the sender rate, then builds the sender rate back up. Additionally, if an
ACK packet in the reverse direction is lost, TCP packets will be
retransmitted (even if
sucessfully received the first time).
In order to disinguish the effective frame throughput from the raw frame
throughput, the term 'goodput' was coined. Throughput could have many
different meanings, thus leading to confusion.
-- Tom
"James K." <txdiversity@hotmail.com> wrote in message
news:bqacjb$42f$1@news1.kornet.net...
> I don't catch what really mean of goodput?
> What is different from throughput?
> Let me know any goodmark about it if you have gotten!
>
> --
> Best regards,
> James K. (txdiversity@hotmail.com, http://home.naver.com/txdiversity)
> - Any remarks, proposal and/or indicator to text would be greatly
respected.
> - Private opinions: These are not the opinions from my affiliation.
>
>
Reply by ●November 30, 20032003-11-30
"TOM" <noname@noprovider.nodomain> wrote in message news:thnyb.12172$a54.1791@nwrddc03.gnilink.net...> Goodput is used to describe the actual throughput of TCP packets, taking > into > account the behavior of TCP backoff, slow start, congestion, and duplicate > packets. > > If a packet is lost, TCP interprets the loss as congestion, and throttles > back > the sender rate, then builds the sender rate back up. Additionally, if an > ACK packet in the reverse direction is lost, TCP packets will be > retransmitted (even if > sucessfully received the first time). > > In order to disinguish the effective frame throughput from the raw frame > throughput, the term 'goodput' was coined. Throughput could have many > different meanings, thus leading to confusion.Tom, Thanks. Yecchhh!! How unfortunate. In my mind, throughput is throughput. Let's see: Let's always start by defining the flow through a frictionless pipe as throughflow. Then let's define the flow through a real pipe with friction as goodflow. Let's always start by defining the current through an ideal wire as current. Then let's define the flow through a real wire with resistance as goodcurrent. ewwwwww!!! Fred
Reply by ●November 30, 20032003-11-30
On Sun, 30 Nov 2003 14:38:17 GMT, "TOM" <noname@noprovider.nodomain> wrote:>Goodput is used to describe the actual throughput of TCP packets, taking >into >account the behavior of TCP backoff, slow start, congestion, and duplicate >packets. > >If a packet is lost, TCP interprets the loss as congestion, and throttles >back >the sender rate, then builds the sender rate back up. Additionally, if an >ACK packet in the reverse direction is lost, TCP packets will be >retransmitted (even if >sucessfully received the first time). > >In order to disinguish the effective frame throughput from the raw frame >throughput, the term 'goodput' was coined. Throughput could have many >different meanings, thus leading to confusion. > > -- Tom >Hi Tom, if you ever find the knucklehead who coined the term "goodput", would you please punch him in the nose for me? Thanks, [-Rick-]
Reply by ●November 30, 20032003-11-30
Rick Lyons wrote:> On Sun, 30 Nov 2003 14:38:17 GMT, "TOM" <noname@noprovider.nodomain> > wrote: > > >>Goodput is used to describe the actual throughput of TCP packets, taking >>into >>account the behavior of TCP backoff, slow start, congestion, and duplicate >>packets. >> >>If a packet is lost, TCP interprets the loss as congestion, and throttles >>back >>the sender rate, then builds the sender rate back up. Additionally, if an >>ACK packet in the reverse direction is lost, TCP packets will be >>retransmitted (even if >>sucessfully received the first time). >> >>In order to disinguish the effective frame throughput from the raw frame >>throughput, the term 'goodput' was coined. Throughput could have many >>different meanings, thus leading to confusion. >> >> -- Tom >> > > > Hi Tom, > if you ever find the knucklehead who coined the > term "goodput", would you please punch him in the > nose for me? > > Thanks, > [-Rick-]I had a hunch about the meaning. After all, we need some word for the concept. When piloting a boat, we use "ground made good" to account for the effects of wind, tide, and current. With GPS, that's direct measurement. Without, it's a distillation of measurement and educated (one hopes) guess. Are stop and stop bits "overhead"? Are headers? I'm tempted to call everything sent "putput", but then we're back to boats. :-) As a name for the useful stuff received, how about "effective data rate"? Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●November 30, 20032003-11-30
r.lyons@REMOVE.ieee.org (Rick Lyons) writes:> if you ever find the knucklehead who coined the > term "goodput", would you please punch him in the > nose for me?Doubleplusgood!!! Ciao, Peter K. -- Peter J. Kootsookos "I will ignore all ideas for new works [..], the invention of which has reached its limits and for whose improvement I see no further hope." - Julius Frontinus, c. AD 84
Reply by ●December 1, 20032003-12-01
"Fred Marshall" <fmarshallx@remove_the_x.acm.org> wrote in message news:<B4CdnbJATOVcxFei4p2dnA@centurytel.net>...> Let's always start by defining the current through an ideal wire as current. > Then let's define the flow through a real wire with resistance as > goodcurrent. > > ewwwwww!!! > > FredHi Fred. Just to add to your agony, what you suggest here is actually used in practical systems (although in the context of power, not current). I used to work on a metal plant between semesters when I went to university. The furnaces worked with AC electrical power, and the various transformers, electrical arcs and other subsystems represented quite substantial reactive loads. On our furnace monitoring systems, there were two displays of power: The "Mega Watt number" showed the actual "good" power consumtion in the furnace, i.e. the power that actually was transformed to heat in the system, and the "Mega VAr number" that represented rective loads internal to the electrical supply system. None of the regular workers could explain to me what that "MVAr number" meant, except it was due to "something electrical that messed with the power supply system". I only learned about reactive loads well into college. I can't remember the reactive measurement actually being used for anything, but the numbers were approximately equal. A furnace running at 40 MW efficient power would usually run at ~35-40 MVAr reactive power. So "the MW number" was the "goodpower" and "the MVAr number" was the "throughpower". I suspect it may be worth being aware of such parameters. The reactive load means way more current flowing through the system than the effective load alone indicates. Rune
Reply by ●December 1, 20032003-12-01
Jerry Avins <jya@ieee.org> wrote in message news:<bqe6nt$grt$1@bob.news.rcn.net>...> Rick Lyons wrote: > > > On Sun, 30 Nov 2003 14:38:17 GMT, "TOM" <noname@noprovider.nodomain> > > wrote: > > > > > >>Goodput is used to describe the actual throughput of TCP packets, taking > >>into > >>account the behavior of TCP backoff, slow start, congestion, and duplicate > >>packets. > >> > >>If a packet is lost, TCP interprets the loss as congestion, and throttles > >>back > >>the sender rate, then builds the sender rate back up. Additionally, if an > >>ACK packet in the reverse direction is lost, TCP packets will be > >>retransmitted (even if > >>sucessfully received the first time). > >> > >>In order to disinguish the effective frame throughput from the raw frame > >>throughput, the term 'goodput' was coined. Throughput could have many > >>different meanings, thus leading to confusion. > >> > >> -- Tom > >> > > > > > > Hi Tom, > > if you ever find the knucklehead who coined the > > term "goodput", would you please punch him in the > > nose for me? > > > > Thanks, > > [-Rick-] > > I had a hunch about the meaning. After all, we need some word for the > concept. When piloting a boat, we use "ground made good" to account for > the effects of wind, tide, and current. With GPS, that's direct > measurement. Without, it's a distillation of measurement and educated > (one hopes) guess. > > Are stop and stop bits "overhead"? Are headers? I'm tempted to call > everything sent "putput", but then we're back to boats. :-) As a name > for the useful stuff received, how about "effective data rate"? > > JerryEh... isn't this a variant of the bit rate/baud rate thing in communications? If I remember correctrly, the bit rate says how much binary data goes through a channel while the baud rate says how much information ((bit rate minus overhead) divided by bits/symbol) goes through? Rune
