Hi all, Is therany onknow the command to generate bursts of packet loss in Netem? Pleashelp to providthe syntax of this command. Thanks, Phuong Dinh Froshemminger aosdl.org Wed Jun 7 12:34:48 2006 From: shemminger aosdl.org (Stephen Hemminger) Date: Wed Apr 18 12:51:19 2007 Subject: Bursts of packeloss In-Reply-To: <D65D93A56024564097FD93238311660A0968F17A@xxxxxxxxxxxxxxxxxxxxxxxx> References: <20060515102133.0353185e@localhost.localdomain> <D65D93A56024564097FD93238311660A0968F17A@xxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <20060607123448.1577c9ea@localhost.localdomain> OWed, 7 Jun 2006 14:25:13 -0400 "dtnphuong" <dtnphuong@xxxxxxxxxx> wrote: > Hi all, > > Is therany onknow the command to generate bursts of packet loss in > Netem? > Pleashelp to providthe syntax of this command. > > Thanks, > Phuong Dinh If you hava loss percentagwith a high correlation, then you should sebursty loss. Nosure though. Frodoom2quakat gmail.com Thu Jun 8 14:45:26 2006 From: doom2quakagmail.com (Dipankar Sarkar) Date: Wed Apr 18 12:51:19 2007 Subject: stautility Message-ID: <3cb7dd3f0606081445j485094fdo8842c7e93e816322@xxxxxxxxxxxxxx> Hi I acurrently trying to emulata network using netem , and was trying to creata distribution tablfor the network latency portion. So, I ra"maketable" to gethe distribution table, now where do i gethstat utility to calculate the rest of the values .. FROM THE DOCUMENTATION ----- 4. Thtables produced arall normalized for mean 0 and standard deviatio1. How do you know whavalues to use for real? Here, I've provided a simpl"stats" utility. Givit a series of floating point values, and iwill return their mean (mu), standard deviation (sigma), and correlatiocoefficien(rho). You can then plug these values directly into NIST Net. ------ canfigurout the stats part ... Thanks iadvanc... Dipankar Sarkar Under graduatstudent Depof computer sciencand eng. IIT Delhi, India Frodtnphuong atma.com.vn Fri Jun 9 10:22:08 2006 From: dtnphuong atma.com.vn (dtnphuong) Date: Wed Apr 18 12:51:19 2007 Subject: ] Install Bridgand Netein Red Hat 9 In-Reply-To: <20060607123448.1577c9ea@localhost.localdomain> Message-ID: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> Hi Stephen, I'vgoa problem in Netem installation with bridge-utils-1.0.6 in Red Hat 9, kernel 2.4.20-8. Thattachmenis the log file for the installation. Pleashelp to taka look on it and help me to solve this problem. Thank you very much, Phuong Dinh -------------- nexpar-------------- A non-texattachmenwas scrubbed... Name: logfile.doc Type: application/msword Size: 32256 bytes Desc: noavailable Url : http://lists.linux-foundation.org/pipermail/netem/attachments/20060609/f74f78fd/logfile.doc Froneteat smyles.plus.com Fri Jun 9 11:54:31 2006 From: neteasmyles.plus.com (Barry Myles) Date: Wed Apr 18 12:51:19 2007 Subject: ] Install Bridgand Netein Red Hat 9 In-Reply-To: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> References: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <4489C3E7.6020908@xxxxxxxxxxxxxxx> dtnphuong wrote: > Thattachmenis the log file for the installation. > Pleashelp to taka look on it and help me to solve this problem. Wherdoes this shell scrip'bridge.sh' come from? Is it something you'vwritten yourself? Could you do thfollowing: - Posthfile 'bridge.sh' to the mailing list. We need to check that th#! linis correct at the start of the file. This could give the error you'rseeing. - Posthoutput of the command 'file bridge.sh'. We need to check that thfilhas unix style rather than windows style line feeds. This too would also givtherror you've seen. - Posthoutput of 'ls -l' . We need to check that the 'x' bits are correctly seon thfile. Just for completeness. Posting a MicrosofWord documento a linux-based mailing list isn't a greaidea - there'll bquite a few people who can't or won't read it. -- Barry Myles Froshemminger aosdl.org Fri Jun 9 14:17:57 2006 From: shemminger aosdl.org (Stephen Hemminger) Date: Wed Apr 18 12:51:19 2007 Subject: ] Install Bridgand Netein Red Hat 9 In-Reply-To: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> References: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <4489E585.3030604@xxxxxxxx> dtnphuong wrote: > Hi Stephen, > > I'vgoa problem in Netem installation with bridge-utils-1.0.6 in Red Hat > 9, kernel 2.4.20-8. > Thattachmenis the log file for the installation. > Pleashelp to taka look on it and help me to solve this problem. > > Thank you very much, > Phuong Dinh > Thakernel (2.4.20) is too old to providnetem. You can make netem work i2.4, buit needs a more recent kernel and the timer resolution (100 HZ) limits thusability. You would bmuch better to try a morrecent distribution (free) like Fedora Coror Ubuntu thasupports the 2.6 kernel. Froshemminger aosdl.org Thu Jun 15 16:24:35 2006 From: shemminger aosdl.org (Stephen Hemminger) Date: Wed Apr 18 12:51:19 2007 Subject: stautility In-Reply-To: <3cb7dd3f0606081445j485094fdo8842c7e93e816322@xxxxxxxxxxxxxx> References: <3cb7dd3f0606081445j485094fdo8842c7e93e816322@xxxxxxxxxxxxxx> Message-ID: <20060615162435.5c7fd821@xxxxxxxxxxxxxxxxx> OFri, 9 Jun 2006 03:15:26 +0530 "Dipankar Sarkar" <doom2quake@xxxxxxxxx> wrote: > Hi > > I acurrently trying to emulata network using netem , and was > trying to creata distribution tablfor the network latency portion. > So, I ra"maketable" to gethe distribution table, now where do i > gethstat utility to calculate the rest of the values .. To usyour table, puit in /usr/lib/tc/XXXX.dist > FROM THE DOCUMENTATION > ----- > 4. Thtables produced arall normalized for mean 0 and standard > deviatio1. How do you know whavalues to use for real? Here, I've > provided a simpl"stats" utility. Givit a series of floating point > values, and iwill return their mean (mu), standard deviation (sigma), > and correlatiocoefficien(rho). You can then plug these values > directly into NIST Net. I didn'copy thaover from NIST Net. I put it into the iproute2 git repository for thnexversion. git://git.kernel.org/pub/scm/linux/kernel/shemminger/iproute2.git or jusbuild thattached file. -------------- nexpar-------------- A non-texattachmenwas scrubbed... Name: stats.c Type: text/x-csrc Size: 1562 bytes Desc: noavailable Url : http://lists.linux-foundation.org/pipermail/netem/attachments/20060615/e6e4e7a7/stats.c Frodoom2quakat gmail.com Mon Jun 26 15:17:32 2006 From: doom2quakagmail.com (Dipankar Sarkar) Date: Wed Apr 18 12:51:19 2007 Subject: ratcontrol example Message-ID: <3cb7dd3f0606261517o6b7a8f86ved451c9c8b718a6e@xxxxxxxxxxxxxx> Hi I required somhelp for ratcontrol .. the example given is not working isays "invalid argument" whethtbf is registered # tc qdisc add dev eth0 roohandl1:0 netem delay 100ms This works fine # tc qdisc add dev eth0 paren1:1 handl10: tbf rate 256kbit buffer 1600 limi3000 RTNETLINK answers: Invalid argument I dnknow how isolve it ... could u help me out on this one ... plus, artherany better ways of rate control ... regards, Dipankar Frodtnphuong atma.com.vn Wed Jun 7 11:25:13 2006 From: dtnphuong atma.com.vn (dtnphuong) Date: Wed Apr 18 17:37:49 2007 Subject: Bursts of packeloss In-Reply-To: <20060515102133.0353185e@localhost.localdomain> Message-ID: <D65D93A56024564097FD93238311660A0968F17A@xxxxxxxxxxxxxxxxxxxxxxxx> Hi all, Is therany onknow the command to generate bursts of packet loss in Netem? Pleashelp to providthe syntax of this command. Thanks, Phuong Dinh Froshemminger aosdl.org Wed Jun 7 12:34:48 2006 From: shemminger aosdl.org (Stephen Hemminger) Date: Wed Apr 18 17:37:49 2007 Subject: Bursts of packeloss In-Reply-To: <D65D93A56024564097FD93238311660A0968F17A@xxxxxxxxxxxxxxxxxxxxxxxx> References: <20060515102133.0353185e@localhost.localdomain> <D65D93A56024564097FD93238311660A0968F17A@xxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <20060607123448.1577c9ea@localhost.localdomain> OWed, 7 Jun 2006 14:25:13 -0400 "dtnphuong" <dtnphuong@xxxxxxxxxx> wrote: > Hi all, > > Is therany onknow the command to generate bursts of packet loss in > Netem? > Pleashelp to providthe syntax of this command. > > Thanks, > Phuong Dinh If you hava loss percentagwith a high correlation, then you should sebursty loss. Nosure though. Frodoom2quakat gmail.com Thu Jun 8 14:45:26 2006 From: doom2quakagmail.com (Dipankar Sarkar) Date: Wed Apr 18 17:37:49 2007 Subject: stautility Message-ID: <3cb7dd3f0606081445j485094fdo8842c7e93e816322@xxxxxxxxxxxxxx> Hi I acurrently trying to emulata network using netem , and was trying to creata distribution tablfor the network latency portion. So, I ra"maketable" to gethe distribution table, now where do i gethstat utility to calculate the rest of the values .. FROM THE DOCUMENTATION ----- 4. Thtables produced arall normalized for mean 0 and standard deviatio1. How do you know whavalues to use for real? Here, I've provided a simpl"stats" utility. Givit a series of floating point values, and iwill return their mean (mu), standard deviation (sigma), and correlatiocoefficien(rho). You can then plug these values directly into NIST Net. ------ canfigurout the stats part ... Thanks iadvanc... Dipankar Sarkar Under graduatstudent Depof computer sciencand eng. IIT Delhi, India Frodtnphuong atma.com.vn Fri Jun 9 10:22:08 2006 From: dtnphuong atma.com.vn (dtnphuong) Date: Wed Apr 18 17:37:49 2007 Subject: ] Install Bridgand Netein Red Hat 9 In-Reply-To: <20060607123448.1577c9ea@localhost.localdomain> Message-ID: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> Hi Stephen, I'vgoa problem in Netem installation with bridge-utils-1.0.6 in Red Hat 9, kernel 2.4.20-8. Thattachmenis the log file for the installation. Pleashelp to taka look on it and help me to solve this problem. Thank you very much, Phuong Dinh -------------- nexpar-------------- A non-texattachmenwas scrubbed... Name: logfile.doc Type: application/msword Size: 32256 bytes Desc: noavailable Url : http://lists.linux-foundation.org/pipermail/netem/attachments/20060609/f74f78fd/logfile-0001.doc Froneteat smyles.plus.com Fri Jun 9 11:54:31 2006 From: neteasmyles.plus.com (Barry Myles) Date: Wed Apr 18 17:37:49 2007 Subject: ] Install Bridgand Netein Red Hat 9 In-Reply-To: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> References: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <4489C3E7.6020908@xxxxxxxxxxxxxxx> dtnphuong wrote: > Thattachmenis the log file for the installation. > Pleashelp to taka look on it and help me to solve this problem. Wherdoes this shell scrip'bridge.sh' come from? Is it something you'vwritten yourself? Could you do thfollowing: - Posthfile 'bridge.sh' to the mailing list. We need to check that th#! linis correct at the start of the file. This could give the error you'rseeing. - Posthoutput of the command 'file bridge.sh'. We need to check that thfilhas unix style rather than windows style line feeds. This too would also givtherror you've seen. - Posthoutput of 'ls -l' . We need to check that the 'x' bits are correctly seon thfile. Just for completeness. Posting a MicrosofWord documento a linux-based mailing list isn't a greaidea - there'll bquite a few people who can't or won't read it. -- Barry Myles Froshemminger aosdl.org Fri Jun 9 14:17:57 2006 From: shemminger aosdl.org (Stephen Hemminger) Date: Wed Apr 18 17:37:49 2007 Subject: ] Install Bridgand Netein Red Hat 9 In-Reply-To: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> References: <D65D93A56024564097FD93238311660A0968F1F7@xxxxxxxxxxxxxxxxxxxxxxxx> Message-ID: <4489E585.3030604@xxxxxxxx> dtnphuong wrote: > Hi Stephen, > > I'vgoa problem in Netem installation with bridge-utils-1.0.6 in Red Hat > 9, kernel 2.4.20-8. > Thattachmenis the log file for the installation. > Pleashelp to taka look on it and help me to solve this problem. > > Thank you very much, > Phuong Dinh > Thakernel (2.4.20) is too old to providnetem. You can make netem work i2.4, buit needs a more recent kernel and the timer resolution (100 HZ) limits thusability. You would bmuch better to try a morrecent distribution (free) like Fedora Coror Ubuntu thasupports the 2.6 kernel. Froshemminger aosdl.org Thu Jun 15 16:24:35 2006 From: shemminger aosdl.org (Stephen Hemminger) Date: Wed Apr 18 17:37:49 2007 Subject: stautility In-Reply-To: <3cb7dd3f0606081445j485094fdo8842c7e93e816322@xxxxxxxxxxxxxx> References: <3cb7dd3f0606081445j485094fdo8842c7e93e816322@xxxxxxxxxxxxxx> Message-ID: <20060615162435.5c7fd821@xxxxxxxxxxxxxxxxx> OFri, 9 Jun 2006 03:15:26 +0530 "Dipankar Sarkar" <doom2quake@xxxxxxxxx> wrote: > Hi > > I acurrently trying to emulata network using netem , and was > trying to creata distribution tablfor the network latency portion. > So, I ra"maketable" to gethe distribution table, now where do i > gethstat utility to calculate the rest of the values .. To usyour table, puit in /usr/lib/tc/XXXX.dist > FROM THE DOCUMENTATION > ----- > 4. Thtables produced arall normalized for mean 0 and standard > deviatio1. How do you know whavalues to use for real? Here, I've > provided a simpl"stats" utility. Givit a series of floating point > values, and iwill return their mean (mu), standard deviation (sigma), > and correlatiocoefficien(rho). You can then plug these values > directly into NIST Net. I didn'copy thaover from NIST Net. I put it into the iproute2 git repository for thnexversion. git://git.kernel.org/pub/scm/linux/kernel/shemminger/iproute2.git or jusbuild thattached file. -------------- nexpar-------------- A non-texattachmenwas scrubbed... Name: stats.c Type: text/x-csrc Size: 1562 bytes Desc: noavailable Url : http://lists.linux-foundation.org/pipermail/netem/attachments/20060615/e6e4e7a7/stats-0001.c Frodoom2quakat gmail.com Mon Jun 26 15:17:32 2006 From: doom2quakagmail.com (Dipankar Sarkar) Date: Wed Apr 18 17:37:49 2007 Subject: ratcontrol example Message-ID: <3cb7dd3f0606261517o6b7a8f86ved451c9c8b718a6e@xxxxxxxxxxxxxx> Hi I required somhelp for ratcontrol .. the example given is not working isays "invalid argument" whethtbf is registered # tc qdisc add dev eth0 roohandl1:0 netem delay 100ms This works fine # tc qdisc add dev eth0 paren1:1 handl10: tbf rate 256kbit buffer 1600 limi3000 RTNETLINK answers: Invalid argument I dnknow how isolve it ... could u help me out on this one ... plus, artherany better ways of rate control ... regards, Dipankar
