Hi ? ? ? I ausing Netein a bridge scenario. I have added loss attribute with correlation. I am facing the unpredictable results when i increases correlation with loss.?For example when i have 40% loss with 0% correlation then the packet drop?accurately?40% at receiver ends. but when i increase correlation .i.e. 50% and 90%, then the?overall?packet drop at receiver end decreases to 30% and 7% ?respectively. Pleashelp. ? Kind Regards Faraz ? Regards Faraz -------------- nexpar-------------- AHTML attachmenwas scrubbed... URL: <http://lists.linuxfoundation.org/pipermail/netem/attachments/20130503/7dc5a874/attachment.html> Froosenbach aus.ibm.com Fri May 3 16:04:02 2013 From: osenbach aus.ibm.co(Bryan Osenbach) Date: Fri, 3 May 2013 10:04:02 -0600 Subject: AUTO: BryaOsenbach is ouof the office (returning 05/08/2013) Message-ID: <OFD1E77F0C.799FC895-ON87257B60.005842BA-87257B60.005842BA@xxxxxxxxxx> I aouof the office until 05/08/2013. Note: This is aautomated responsto your message "Netem Digest, Vol 81, Issu1" senon 05/03/2013 6:00:04. This is thonly notification you will receivwhile this person is away. -------------- nexpar-------------- AHTML attachmenwas scrubbed... URL: <http://lists.linuxfoundation.org/pipermail/netem/attachments/20130503/1b1ce28e/attachment.html> Frofaraz4557 ayahoo.com Sat May 4 08:44:37 2013 From: faraz4557 ayahoo.co(faraz iqbal) Date: Sat, 4 May 2013 01:44:37 -0700 (PDT) Subject: NeteCorrelation In-Reply-To: <20130503073639.18897ddc@xxxxxxxxxxxxxxxxxxxxxxxxxxx> References: <1902752B0C92F943AB7EA9EE13E2DEEC0DE53328E9@xxxxxxxxxxxxxxxxxxxxxxxxxxx> <20130503073639.18897ddc@xxxxxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <1367657077.77526.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> How cai usnew loss state and gemodel. Kindly suggest. ? Regards Faraz ________________________________ From: StepheHemminger <stephen anetworkplumber.org> To: faraz iqbal <faraz4557 ayahoo.com> Sent: Friday, May 3, 2013 4:36 PM Subject: Re: NeteCorrelation OFri, 3 May 2013 03:34:41 -0700 faraz iqbal <faraz4557 ayahoo.com> wrote: > Hi >? ? ? I ausing Netein a bridge scenario. I have added loss attribute with correlation. I am facing the unpredictable results when i increases correlation with loss. For example when i have 40% loss with 0% correlation then the packet drop accurately 40% at receiver ends. but when i increase correlation .i.e. 50% and 90%, then the overall packet drop at receiver end decreases to 30% and 7%? respectively. > > Pleashelp. > > Kind Regards > > Faraz > > Regards > > Faraz Thold loss model is really rough. Usone of the new loss state or gemodel instead. -------------- nexpar-------------- AHTML attachmenwas scrubbed... URL: <http://lists.linuxfoundation.org/pipermail/netem/attachments/20130504/de3ebfbd/attachment.html> Frostefano.salsano auniroma2.it Sat May 4 13:19:36 2013 From: stefano.salsano auniroma2.i(Stefano Salsano) Date: Sat, 04 May 2013 15:19:36 +0200 Subject: NeteCorrelation In-Reply-To: <1367657077.77526.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> References: <1902752B0C92F943AB7EA9EE13E2DEEC0DE53328E9@xxxxxxxxxxxxxxxxxxxxxxxxxxx> <20130503073639.18897ddc@xxxxxxxxxxxxxxxxxxxxxxxxxxx> <1367657077.77526.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <51850AE8.6000208@xxxxxxxxxxx> Il 04/05/2013 10:44, faraz iqbal ha scritto: > How cai usnew loss state and gemodel. Kindly suggest. you may hava look herhttp://netgroup.uniroma2.it/NetemCLG/ regards Stefano > Faraz > ------------------------------------------------------------------------ > *From:* StepheHemminger <stephen anetworkplumber.org> > OFri, 3 May 2013 03:34:41 -0700 > faraz iqbal <faraz4557 ayahoo.co<mailto:faraz4557 at yahoo.com>> wrote: > I ausing Netein a bridge scenario. I have added loss > attributwith correlation. I afacing the unpredictable results when i > increases correlatiowith loss. For examplwhen i have 40% loss with > 0% correlatiothen thpacket drop accurately 40% at receiver ends. but > whei increascorrelation .i.e. 50% and 90%, then the overall packet > drop areceiver end decreases to 30% and 7% respectively. > > > > Thold loss model is really rough. Usone of the new loss state or > gemodel instead. > > > > > _______________________________________________ > Netemailing list > Netealists.linux-foundation.org > https://lists.linuxfoundation.org/mailman/listinfo/netem > -- ******************************************************************* Stefano Salsano Dipartimento Ingegneria Elettronica Universita' di Roma "Tor Vergata" Via del Politecnico, 1 - 00133 Roma - ITALY http://netgroup.uniroma2.it/Stefano_Salsano/ E-mail : stefano.salsano auniroma2.it Cell. : +39 320 4307310 Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 ******************************************************************* Frofaraz4557 ayahoo.com Wed May 8 11:35:36 2013 From: faraz4557 ayahoo.co(faraz iqbal) Date: Wed, 8 May 2013 04:35:36 -0700 (PDT) Subject: Neteloss Message-ID: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Hi I awilling to add loss in burspattern. How i can set values in gemodel (p, r, 1-h & 1-k).? I am little bit confuse, have some one give an example. ? Kind Regards Faraz -------------- nexpar-------------- AHTML attachmenwas scrubbed... URL: <http://lists.linuxfoundation.org/pipermail/netem/attachments/20130508/d352dc3b/attachment.html> Frostefano.salsano auniroma2.it Wed May 8 12:33:12 2013 From: stefano.salsano auniroma2.i(Stefano Salsano) Date: Wed, 08 May 2013 14:33:12 +0200 Subject: Neteloss In-Reply-To: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> References: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <518A4608.7090504@xxxxxxxxxxx> Il 08/05/2013 13:35, faraz iqbal ha scritto: > Hi > I awilling to add loss in burspattern. How i can set values in > gemodel (p, r, 1-h & 1-k).? I alittlbit confuse, have some one give > aexample. > Thsimplesthing you can do is to simulate a loss patterns in which packets arlosin bursts of consecutive loss events ithis casyou only need to set p and r sepag13 of http://netgroup.uniroma2.it/twiki/pub/Main/NetemCLG/TR-loss-netem.pdf Th?SimplGilbert? model has two states and two independent parameters, p and r. Iis ablto model a system with consecutive loss events, which cabcharacterized by a loss probability and a bursduration. for example tc qdisc add dev eth0 rooneteloss gemodel 5 50 will setup a loss patterwherthe average burst duration is E(B) = 1/r = 2 and thloss probability is ploss = p/(p+r) = 0.05/0.55 = 1/11 (sethformulas on page 30 of http://netgroup.uniroma2.it/twiki/pub/Main/NetemCLG/TR-loss-netem.pdf) > Kind Regards > > Faraz > > > > _______________________________________________ > Netemailing list > Netealists.linux-foundation.org > https://lists.linuxfoundation.org/mailman/listinfo/netem > -- ******************************************************************* Stefano Salsano Dipartimento Ingegneria Elettronica Universita' di Roma "Tor Vergata" Via del Politecnico, 1 - 00133 Roma - ITALY http://netgroup.uniroma2.it/Stefano_Salsano/ E-mail : stefano.salsano auniroma2.it Cell. : +39 320 4307310 Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 ******************************************************************* Frofaraz4557 ayahoo.com Wed May 8 15:09:45 2013 From: faraz4557 ayahoo.co(faraz iqbal) Date: Wed, 8 May 2013 08:09:45 -0700 (PDT) Subject: Neteloss In-Reply-To: <518A4608.7090504@xxxxxxxxxxx> References: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A4608.7090504@xxxxxxxxxxx> Message-ID: <1368025785.75810.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> How i calocatnetem-patch catalog in my computer where i am running netem.? ?I havlinux 3.0.32. ? Regards Faraz ________________________________ From: Stefano Salsano <stefano.salsano auniroma2.it> To: netealists.linux-foundation.org Sent: Wednesday, May 8, 2013 2:33 PM Subject: Re: Neteloss Il 08/05/2013 13:35, faraz iqbal ha scritto: > Hi > I awilling to add loss in burspattern. How i can set values in > gemodel (p, r, 1-h & 1-k).? I alittlbit confuse, have some one give > aexample. > Thsimplesthing you can do is to simulate a loss patterns in which packets arlosin bursts of consecutive loss events ithis casyou only need to set p and r sepag13 of http://netgroup.uniroma2.it/twiki/pub/Main/NetemCLG/TR-loss-netem.pdf Th?SimplGilbert? model has two states and two independent parameters, p and r. Iis ablto model a system with consecutive loss events, which cabcharacterized by a loss probability and a bursduration. for example tc qdisc add dev eth0 rooneteloss gemodel 5 50 will setup a loss patterwherthe average burst duration is E(B) = 1/r = 2 and thloss probability is ploss = p/(p+r) = 0.05/0.55 = 1/11 (sethformulas on page 30 of http://netgroup.uniroma2.it/twiki/pub/Main/NetemCLG/TR-loss-netem.pdf) > Kind Regards > > Faraz > > > > _______________________________________________ > Netemailing list > Netealists.linux-foundation.org > https://lists.linuxfoundation.org/mailman/listinfo/netem > -- ******************************************************************* Stefano Salsano Dipartimento Ingegneria Elettronica Universita' di Roma "Tor Vergata" Via del Politecnico, 1 - 00133 Roma - ITALY http://netgroup.uniroma2.it/Stefano_Salsano/ E-mail? : stefano.salsano auniroma2.it Cell.? : +39 320 4307310 Office? : (Tel.) +39 06 72597770? (Fax.) +39 06 72597435 ******************************************************************* _______________________________________________ Netemailing list Netealists.linux-foundation.org https://lists.linuxfoundation.org/mailman/listinfo/netem -------------- nexpar-------------- AHTML attachmenwas scrubbed... URL: <http://lists.linuxfoundation.org/pipermail/netem/attachments/20130508/a3926b85/attachment.html> Frostefano.salsano auniroma2.it Wed May 8 15:32:07 2013 From: stefano.salsano auniroma2.i(Stefano Salsano) Date: Wed, 08 May 2013 17:32:07 +0200 Subject: Neteloss In-Reply-To: <1368025785.75810.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> References: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A4608.7090504@xxxxxxxxxxx> <1368025785.75810.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <518A6FF7.7050005@xxxxxxxxxxx> Il 08/05/2013 17:09, faraz iqbal ha scritto: > How i calocatnetem-patch catalog in my computer where i am running > netem.? > > I havlinux 3.0.32. if I'nowrong, the needed code is already integrated in the kernel 3.0 thesarthe kernel source files: http://lxr.free-electrons.com/source/net/sched/sch_netem.c?v=3.0 http://lxr.free-electrons.com/source/include/linux/pkt_sched.h?v=3.0 iiproute2 threlevant source files are: iproute2-3.1.0\tc\q_netem.c iproute2-3.1.0\include\linux\pkt_sched.h > Regards > > Faraz > ------------------------------------------------------------------------ > *From:* Stefano Salsano <stefano.salsano auniroma2.it> > *To:* netealists.linux-foundation.org > *Sent:* Wednesday, May 8, 2013 2:33 PM > *Subject:* Re: Neteloss > > Il 08/05/2013 13:35, faraz iqbal ha scritto: > > Hi > > I awilling to add loss in burspattern. How i can set values in > > gemodel (p, r, 1-h & 1-k).? I alittlbit confuse, have some one give > > aexample. > > > > Thsimplesthing you can do is to simulate a loss patterns in which > packets arlosin bursts of consecutive loss events > > ithis casyou only need to set p and r > > sepag13 of > http://netgroup.uniroma2.it/twiki/pub/Main/NetemCLG/TR-loss-netem.pdf > > Th?SimplGilbert? model has two states and two independent > parameters, p and r. Iis ablto model a system with consecutive loss > events, which cabcharacterized by a loss probability and a > bursduration. > > for example > tc qdisc add dev eth0 rooneteloss gemodel 5 50 > > will setup a loss patterwherthe average burst duration is > E(B) = 1/r = 2 > > and thloss probability is > ploss = p/(p+r) = 0.05/0.55 = 1/11 > > (sethformulas on page 30 of > http://netgroup.uniroma2.it/twiki/pub/Main/NetemCLG/TR-loss-netem.pdf) > > > > > Kind Regards > > > > Faraz > > > > > > > > _______________________________________________ > > Netemailing list > > Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> > > https://lists.linuxfoundation.org/mailman/listinfo/netem > > > > > -- > ******************************************************************* > Stefano Salsano > Dipartimento Ingegneria Elettronica > Universita' di Roma "Tor Vergata" > Via del Politecnico, 1 - 00133 Roma - ITALY > > http://netgroup.uniroma2.it/Stefano_Salsano/ > > E-mail : stefano.salsano auniroma2.i<mailto:stefano.salsano at uniroma2.it> > Cell. : +39 320 4307310 > Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 > ******************************************************************* > _______________________________________________ > Netemailing list > Netealists.linux-foundation.org <mailto:Netem at lists.linux-foundation.org> > https://lists.linuxfoundation.org/mailman/listinfo/netem > -- ******************************************************************* Stefano Salsano Dipartimento Ingegneria Elettronica Universita' di Roma "Tor Vergata" Via del Politecnico, 1 - 00133 Roma - ITALY http://netgroup.uniroma2.it/Stefano_Salsano/ E-mail : stefano.salsano auniroma2.it Cell. : +39 320 4307310 Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 ******************************************************************* Frofaraz4557 ayahoo.com Thu May 9 10:22:37 2013 From: faraz4557 ayahoo.co(faraz iqbal) Date: Thu, 9 May 2013 03:22:37 -0700 (PDT) Subject: Neteloss Message-ID: <1368094957.53024.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Hi I hava bridgcase for Netem packet emulation. In case of zero loss condition i generates 5000 packets at my traffic source and i capture entire packets (TX = 5000) at sender ethernet interface. But for emulation case (Netem-loss at receiver end input interface) i have captured less packets (TX = 4950) than my traffic source actuall generated (= 5000).? Is this duto ARP timout or any thing else.? ?Regards Faraz -------------- nexpar-------------- AHTML attachmenwas scrubbed... URL: <http://lists.linuxfoundation.org/pipermail/netem/attachments/20130509/4021e9f9/attachment.html> Frofaraz4557 ayahoo.com Fri May 10 09:36:03 2013 From: faraz4557 ayahoo.co(faraz iqbal) Date: Fri, 10 May 2013 02:36:03 -0700 (PDT) Subject: Neteloss In-Reply-To: <518A6FF7.7050005@xxxxxxxxxxx> References: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A4608.7090504@xxxxxxxxxxx> <1368025785.75810.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A6FF7.7050005@xxxxxxxxxxx> Message-ID: <1368178563.92725.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Hi Stefano ? ? ? ? ? ? ? ? Thanks for your supporand cooperation. You gavme example last time. I have also read ?TR-loss netem report. I have few questions from this example. Imeans thaa single burst duration??E(B)?is 2 sec. Is ploss is thoverall packeloss or it is only packet loss during burst duration? For example i run a test of 1000 seconds. Will ploss is overall packet loss for my 1000 seconds, such as option of loss percentage in netem. ?Iexampl:?ploss= 1/11= 0.09 ?>> Can i say it 9.1 % loss. > example ? > tc qdisc add dev eth0 rooneteloss gemodel 5 50 > > will setup a loss patterwherthe average burst duration is > E(B) = 1/r = 2 > > and thloss probability is > ploss = p/(p+r) = 0.05/0.55 = 1/11 ? ?Kind Regards ?Faraz >? > _______________________________________________ >? > Netemailing list >? > Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> >? > https://lists.linuxfoundation.org/mailman/listinfo/netem >? > > > > -- > ******************************************************************* > Stefano Salsano > Dipartimento Ingegneria Elettronica > Universita' di Roma "Tor Vergata" > Via del Politecnico, 1 - 00133 Roma - ITALY > > http://netgroup.uniroma2.it/Stefano_Salsano/ > > E-mail? : stefano.salsano auniroma2.i<mailto:stefano.salsano at uniroma2.it> > Cell.? : +39 320 4307310 > Office? : (Tel.) +39 06 72597770? (Fax.) +39 06 72597435 > ******************************************************************* > _______________________________________________ > Netemailing list > Netealists.linux-foundation.org <mailto:Netem at lists.linux-foundation.org> > https://lists.linuxfoundation.org/mailman/listinfo/netem > -- ******************************************************************* Stefano Salsano Dipartimento Ingegneria Elettronica Universita' di Roma "Tor Vergata" Via del Politecnico, 1 - 00133 Roma - ITALY http://netgroup.uniroma2.it/Stefano_Salsano/ E-mail? : stefano.salsano auniroma2.it Cell.? : +39 320 4307310 Office? : (Tel.) +39 06 72597770? (Fax.) +39 06 72597435 ******************************************************************* -------------- nexpar-------------- AHTML attachmenwas scrubbed... URL: <http://lists.linuxfoundation.org/pipermail/netem/attachments/20130510/6411e1b0/attachment.html> Frostefano.salsano auniroma2.it Fri May 10 12:17:58 2013 From: stefano.salsano auniroma2.i(Stefano Salsano) Date: Fri, 10 May 2013 14:17:58 +0200 Subject: Neteloss In-Reply-To: <1368178563.92725.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> References: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A4608.7090504@xxxxxxxxxxx> <1368025785.75810.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A6FF7.7050005@xxxxxxxxxxx> <1368178563.92725.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <518CE576.7060609@xxxxxxxxxxx> Il 10/05/2013 11:36, faraz iqbal ha scritto: > Hi Stefano > Thanks for your supporand cooperation. You gavme > exampllastime. I have also read TR-loss netem report. I have few > questions frothis example. > > Imeans thaa single burst duration E(B) is 2 sec. no, E(B) is ipackets. Imeans that the *average* lenght of a burst of lospackets is 2 > Is ploss is thoverall packeloss or it is only packet loss during > bursduration? For exampli run a test of 1000 seconds. Will ploss is ploss is thoverall averagpacket loss during bursyou hav100% loss because you are using the "simple gilbert" model using only two parameters iorder to provida loss probability during bursts you have to provide a third parameter (this is explained ithreport) > overall packeloss for my 1000 seconds, such as option of loss > percentagin netem. > > Iexampl: ploss= 1/11= 0.09 >> Can i say it 9.1 % loss. > >> example > > > tc qdisc add dev eth0 rooneteloss gemodel 5 50 > > > > will setup a loss patterwherthe average burst duration is > > E(B) = 1/r = 2 > > > > and thloss probability is > > ploss = p/(p+r) = 0.05/0.55 = 1/11 > > > > > > Kind Regards > Faraz > > > > _______________________________________________ > > > Netemailing list > > > Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> > > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org>> > > > https://lists.linuxfoundation.org/mailman/listinfo/netem > > > > > > > > > -- > > ******************************************************************* > > Stefano Salsano > > Dipartimento Ingegneria Elettronica > > Universita' di Roma "Tor Vergata" > > Via del Politecnico, 1 - 00133 Roma - ITALY > > > > http://netgroup.uniroma2.it/Stefano_Salsano/ > > > > E-mail : stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it> <mailto:stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it>> > > Cell. : +39 320 4307310 > > Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 > > ******************************************************************* > > _______________________________________________ > > Netemailing list > > Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org>> > > https://lists.linuxfoundation.org/mailman/listinfo/netem > > > > > -- > ******************************************************************* > Stefano Salsano > Dipartimento Ingegneria Elettronica > Universita' di Roma "Tor Vergata" > Via del Politecnico, 1 - 00133 Roma - ITALY > > http://netgroup.uniroma2.it/Stefano_Salsano/ > > E-mail : stefano.salsano auniroma2.i<mailto:stefano.salsano at uniroma2.it> > Cell. : +39 320 4307310 > Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 > ******************************************************************* > > -- ******************************************************************* Stefano Salsano Dipartimento Ingegneria Elettronica Universita' di Roma "Tor Vergata" Via del Politecnico, 1 - 00133 Roma - ITALY http://netgroup.uniroma2.it/Stefano_Salsano/ E-mail : stefano.salsano auniroma2.it Cell. : +39 320 4307310 Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 ******************************************************************* Frofaraz4557 ayahoo.com Sat May 11 09:44:16 2013 From: faraz4557 ayahoo.co(faraz iqbal) Date: Sat, 11 May 2013 02:44:16 -0700 (PDT) Subject: Neteloss In-Reply-To: <518CE576.7060609@xxxxxxxxxxx> References: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A4608.7090504@xxxxxxxxxxx> <1368025785.75810.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A6FF7.7050005@xxxxxxxxxxx> <1368178563.92725.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518CE576.7060609@xxxxxxxxxxx> Message-ID: <1368265456.52950.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Dear Stefano > Imeans thaa single burst duration E(B) is 2 sec. no, E(B) is ipackets. Imeans that the *average* lenght of a burst of lospackets is 2 Faraz : ?Ok, If i anowrong then it means E(B) = number of lost packets in burst.Which is equals to 2 in our example case. during bursyou hav100% loss because you are using the "simple gilbert" model using only two parameters iorder to provida loss probability during bursts you have to provide a third parameter (this is explained ithreport) Faraz : I agreed with third parameter. Bui havstill one confusion which loss that ?"ploss (loss probability)" represents here.??Actually i want to know about overall packet loss. Iexampl: ploss= 1/11= 0.09? >> Can i say it 9.1 % loss. >>? example > >? > tc qdisc add dev eth0 rooneteloss gemodel 5 50 >? > >? > will setup a loss patterwherthe average burst duration is >? > E(B) = 1/r = 2 >? > >? > and thloss probability is >? > ploss = p/(p+r) = 0.05/0.55 = 1/11 > > > > > >? Kind Regards >? Faraz > >? >? > _______________________________________________ >? >? > Netemailing list >? >? > Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> >? > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org>> >? >? > https://lists.linuxfoundation.org/mailman/listinfo/netem >? >? > >? > >? > >? > -- >? > ******************************************************************* >? > Stefano Salsano >? > Dipartimento Ingegneria Elettronica >? > Universita' di Roma "Tor Vergata" >? > Via del Politecnico, 1 - 00133 Roma - ITALY >? > >? > http://netgroup.uniroma2.it/Stefano_Salsano/ >? > >? > E-mail? : stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it> <mailto:stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it>> >? > Cell.? : +39 320 4307310 >? > Office? : (Tel.) +39 06 72597770? (Fax.) +39 06 72597435 >? > ******************************************************************* >? > _______________________________________________ >? > Netemailing list >? > Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org>> >? > https://lists.linuxfoundation.org/mailman/listinfo/netem >? > > > > -- > ******************************************************************* > Stefano Salsano > Dipartimento Ingegneria Elettronica > Universita' di Roma "Tor Vergata" > Via del Politecnico, 1 - 00133 Roma - ITALY > > http://netgroup.uniroma2.it/Stefano_Salsano/ > > E-mail? : stefano.salsano auniroma2.i<mailto:stefano.salsano at uniroma2.it> > Cell.? : +39 320 4307310 > Office? : (Tel.) +39 06 72597770? (Fax.) +39 06 72597435 > ******************************************************************* > > -- ******************************************************************* Stefano Salsano Dipartimento Ingegneria Elettronica Universita' di Roma "Tor Vergata" Via del Politecnico, 1 - 00133 Roma - ITALY http://netgroup.uniroma2.it/Stefano_Salsano/ E-mail? : stefano.salsano auniroma2.it Cell.? : +39 320 4307310 Office? : (Tel.) +39 06 72597770? (Fax.) +39 06 72597435 ******************************************************************* -------------- nexpar-------------- AHTML attachmenwas scrubbed... URL: <http://lists.linuxfoundation.org/pipermail/netem/attachments/20130511/4bbd6d6c/attachment.html> Frostefano.salsano auniroma2.it Mon May 13 12:54:50 2013 From: stefano.salsano auniroma2.i(Stefano Salsano) Date: Mon, 13 May 2013 14:54:50 +0200 Subject: Neteloss In-Reply-To: <1368265456.52950.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> References: <1368012936.61616.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A4608.7090504@xxxxxxxxxxx> <1368025785.75810.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518A6FF7.7050005@xxxxxxxxxxx> <1368178563.92725.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> <518CE576.7060609@xxxxxxxxxxx> <1368265456.52950.YahooMailNeo@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <5190E29A.2070707@xxxxxxxxxxx> Il 11/05/2013 11:44, faraz iqbal ha scritto: > Dear Stefano > > > Imeans thaa single burst duration E(B) is 2 sec. > > no, E(B) is ipackets. Imeans that the *average* lenght of a burst of > lospackets is 2 > > Faraz : Ok, If i anowrong then it means E(B) = number of lost > packets iburst.Which is equals to 2 in our examplcase. E(B) is thnumber of packets in th"burst", in general it does not correspond to thnumber of lospackets in the burst, because in general thloss probability during thburst can be <100% ithexample I gave, "simple gilbert" model, the burst lenght and the number of lospackets in thburst are the same, because the loss probability during thbursis 100% > during bursyou hav100% loss because you are using the "simple > gilbert" model using only two parameters > > iorder to provida loss probability during bursts you have to provide > a third parameter (this is explained ithreport) > > Faraz : I agreed with third parameter. Bui havstill one confusion > which loss tha "ploss (loss probability)" represents here.? Actually i > wanto know abouoverall packet loss. > > Iexampl: ploss= 1/11= 0.09 >> Can i say it 9.1 % loss. > yes, this formula evaluates thoverall packeloss, which is 9.1% loss ithis case > > >> example > > > > > tc qdisc add dev eth0 rooneteloss gemodel 5 50 > > > > > > will setup a loss patterwherthe average burst duration is > > > E(B) = 1/r = 2 > > > > > > and thloss probability is > > > ploss = p/(p+r) = 0.05/0.55 = 1/11 > > > > > > > > > > > > Kind Regards > > Faraz > > > > > > _______________________________________________ > > > > Netemailing list > > > > Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> > > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org>> > > > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> > > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org>>> > > > > https://lists.linuxfoundation.org/mailman/listinfo/netem > > > > > > > > > > > > > -- > > > ******************************************************************* > > > Stefano Salsano > > > Dipartimento Ingegneria Elettronica > > > Universita' di Roma "Tor Vergata" > > > Via del Politecnico, 1 - 00133 Roma - ITALY > > > > > > http://netgroup.uniroma2.it/Stefano_Salsano/ > > > > > > E-mail : stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it> > > <mailto:stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it>> > <mailto:stefano.salsano auniroma2.i<mailto:stefano.salsano at uniroma2.it> > > <mailto:stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it>>> > > > Cell. : +39 320 4307310 > > > Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 > > > ******************************************************************* > > > _______________________________________________ > > > Netemailing list > > > Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> > > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org>> > > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org> > > <mailto:Netealists.linux-foundation.org > <mailto:Netealists.linux-foundation.org>>> > > > https://lists.linuxfoundation.org/mailman/listinfo/netem > > > > > > > > > -- > > ******************************************************************* > > Stefano Salsano > > Dipartimento Ingegneria Elettronica > > Universita' di Roma "Tor Vergata" > > Via del Politecnico, 1 - 00133 Roma - ITALY > > > > http://netgroup.uniroma2.it/Stefano_Salsano/ > > > > E-mail : stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it> <mailto:stefano.salsano auniroma2.it > <mailto:stefano.salsano auniroma2.it>> > > Cell. : +39 320 4307310 > > Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 > > ******************************************************************* > > > > > > > -- > ******************************************************************* > Stefano Salsano > Dipartimento Ingegneria Elettronica > Universita' di Roma "Tor Vergata" > Via del Politecnico, 1 - 00133 Roma - ITALY > > http://netgroup.uniroma2.it/Stefano_Salsano/ > > E-mail : stefano.salsano auniroma2.i<mailto:stefano.salsano at uniroma2.it> > Cell. : +39 320 4307310 > Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 > ******************************************************************* > > -- ******************************************************************* Stefano Salsano Dipartimento Ingegneria Elettronica Universita' di Roma "Tor Vergata" Via del Politecnico, 1 - 00133 Roma - ITALY http://netgroup.uniroma2.it/Stefano_Salsano/ E-mail : stefano.salsano auniroma2.it Cell. : +39 320 4307310 Offic : (Tel.) +39 06 72597770 (Fax.) +39 06 72597435 *******************************************************************
