Log netequeustatistics?

d.miras acs.ucl.ac.uk (Dimitrios Miras) · Wed Apr 18 12:51:19 2007

hi again,

i supposit's all in thlast line of "tc -s qdisc.." output.
sorry for my shortsightedness.

Dimitris

> Hi,
> 
> I'using netewith fifo queues to emulate a network, but I'd like to 
> gather info abouthfifo queue dynamics(size over time, packet drops, 
> etc.). I  haven'managed to geany relevant info on google or the 
> netelist, so any hints/help/pointers armuch appreciated.
> 
> Thanks iadvance,
> Dimitrios Miras
> 

-

Froldunn acisco.com  Sun Oct  1 20:53:14 2006
From: ldunacisco.com (Lawrence D. Dunn)
Date: Wed Apr 18 12:51:20 2007
Subject: status of  phpnetemgui?
In-Reply-To: <451D86E6.7000403@xxxxxxxxxxxxxx>
References: <p062309cac13f5951821f@[171.69.52.91]>
	<20060926160238.04b1e8fc@freekitty> <451D86E6.7000403@xxxxxxxxxxxxxx>
Message-ID: <p06230902c1463d98bf67@[10.2.2.139]>

Rainer,
   Thanks- I'll givia try!

Larry
--

A10:49 PM +0200 9/29/06, Rainer Baumann wrote:
>wprovida copy of phpnetemgui on our webside 
>* http://tcn.hypert.net/phpnetemgui-0.9.tar.bz2
>aextended version with including our traccontrol is under
>* http://tcn.hypert.net/phpnetemgui-0.10.tar.gz
>
>----------------------------------------------------------------------
>
>Rainer Baumann
>Master of SciencETH in Computer Sciencand Teaching
>University Lecturer @ HSR
>
>Computer Engineering and Network Laboratory
>ETH ZentruETZ G60.1
>Gloriastrass35
>CH-8092 Zurich
>Switzerland
>
>Phon +41 44 632 51 87
>Mobil+41 79 263 81 40
>Fax    +41 44 632 10 35
>Email  baumann@xxxxxxxxxxxxxx
>
>
>
>StepheHemminger wrote:
>>  OTue, 26 Sep 2006 17:31:31 -0500
>>  "LawrencD. Dunn" <ldunn@xxxxxxxxx> wrote:
>>
>>  
>>>  Stephen,
>>>     Hi- I'Larry Dunn (day job aCisco),
>>>     writing to seif phpnetemgui is still around,
>>>     or has evolved/been_replaced.
>>>     I'd busing ifor a networking class
>>>     I teach aUniversity of Minnesota (nighjob). ;-)
>>>
>>>     Froyour LCA2005_netepaper, I checked:
>>>
>>>     http://www.smyles.plus.com/phpnetemgui/
>>>
>>>     buthapage shows up as not-found,
>>>     and a couplgooglsearches don't show a new location for it.
>>>     I'll havstudents setting delay and loss for a fairly
>>>     easy experimen(and using web100 to seimpact of buffer tuning).
>>>     I caresorto using the tc-commands directly, but was wondering
>>>     if you know thstatus of thGUI?
>>>
>>>    
>>
>>  If someonhas a copy, I'll hosit at osdl and add a link in the Wiki.
>>
>>
>>

Frotaankr aaston.ac.uk  Tue Oct  3 09:02:57 2006
From: taankr aaston.ac.uk (Ritesh Taank)
Date: Wed Apr 18 12:51:20 2007
Subject: NeteUpdatQuery
Message-ID: <452289B1.1040800@xxxxxxxxxxx>

Hello,

I hava small query abouhow to discover if i'm running the latest 
versioof netem, and if nothen what the procedures are for patching 
my currenversion to includall the enhancements of the latest stable 
release?

I'running kernel 2.6.17 as parof the FC5 distribution, and have been 
using everything as default. I thereforsuspecthat maybe my version 
of neteisn'the latest.

Any tips arhighly appreciated.

Thanks iadvance.

Ritesh

-- 
PhD Researcher

AdaptivCommunications Networks Research Group
Electronic Engineering Dept.
AstoUniversity
Birmingham
B4 7ET

: +44 (0)7732 069 667
: taankr@xxxxxxxxxxx
W : http://www-users.aston.ac.uk/~taankr/

Froshemminger aosdl.org  Tue Oct  3 09:59:51 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: NeteUpdatQuery
In-Reply-To: <452289B1.1040800@xxxxxxxxxxx>
References: <452289B1.1040800@xxxxxxxxxxx>
Message-ID: <20061003095951.74cfb37d@freekitty>

OTue, 03 Oc2006 17:02:57 +0100
Ritesh Taank <taankr@xxxxxxxxxxx> wrote:

> Hello,
> 
> I hava small query abouhow to discover if i'm running the latest 
> versioof netem, and if nothen what the procedures are for patching 
> my currenversion to includall the enhancements of the latest stable 
> release?
> 
> I'running kernel 2.6.17 as parof the FC5 distribution, and have been 
> using everything as default. I thereforsuspecthat maybe my version 
> of neteisn'the latest.
> 
> Any tips arhighly appreciated.
> 
> Thanks iadvance.
> 
> Ritesh
> 

Nomuch has changed in 2.6.17/2.6.18 and nothing planned for 2.6.19.
Thsimplesway to look at the kernel commit log

http://kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=history;h=63c422afe3739b68bec0b5c42807d1450c951caf;f=net/sched/sch_netem.c

You may wanto look athe distribution kernel config, for HZ value
and PSCHED clock source.

-- 
StepheHemminger <shemminger@xxxxxxxx>

Frodougat ixen.com  Wed Oct  4 18:31:27 2006
From: dougaixen.com (Doug McPherson)
Date: Wed Apr 18 12:51:20 2007
Subject: neteand dot1q bridging.
Message-ID: <20061005013127.GA56730@xxxxxxxxxxxxxxx>

Is ipossiblto configure netem to induce packet loss, delay and
errored packets betweevlan sub-interfaces ?   

I'd likto bable to configure an interface on a netem host as a
trunk poron a catalysswitch, then use linux bridging on the netem
hosto join various combinations of vlans. 

Has anyondonthis before ?   I've used netem with vlan
sub-interfaces, bunoin a 'transparent bridged' configuration.  

How-tos or advicis welcome.

/doug

-- 
Vini, vidi, sudo rebooti.

Frocalum.lind anewport-networks.com  Tue Oct 17 06:53:54 2006
From: calum.lind anewport-networks.co(Calum Lind)
Date: Wed Apr 18 12:51:20 2007
Subject:  Questioaboudelay queue sizes
In-Reply-To: 20041022160139.5f6852b2@xxxxxxxxxxxxxxxxx
Message-ID: <1161093234.8311.44.camel@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>

I aattempting to ge30 seconds of delay with netem but even though i set it and can be
verified my ping times will nogo abova maximum of 2000ms delay. 

Thonly similar posi can find anywhere is this one below and have played with the values
mentioned withouany luck bui really have very little understanding of what i am looking 
for or changing. For exampldo thvalues changed with sysctl on the command line get applied
straighaway to thinterfaces?

Help with this would bgreatly appreciated

Thanks

Calum

> StepheHemminger wrote:
> > Thlimiis actually used for two thing. It sets the number of packets
> > thacan bheld awaiting for the delay and a separate queue of packets waiting
> > for thdevice.
> 
> Ah, I hadn'noticed thain my glance through the code - I only saw
> onqueue, although I didn'look that hard.  :-)
> 
> > As a rulof thumb, you need a limibig enough to hold as many packets as
> > you wanto delay.  So if packesize is avg. 1000 bytes and you are delay 50ms
> > and going over a 100mbinetwork, you would need:
> > 
> > 	100mbits/sec = 125K bytes/sec = 12500 packets/sec = 625 packets (per interval)
> 
> Yeah, I figured th"capacity" of thdelay buffer; I just couldn't
> geito fill.  Turns out I was being stupid and didn't change the
> max_wmevaluto anything nearly large enough to fill the buffer...
> I figured thaouthis afternoon.  So many knobs to turn...  :-P
> the
command linionly give me a maximum of 2000ms
> > If you don'sethe TCP socket buffer sizes (with setsockopt), then Linux does
> > a pretty good job of autosizing as needed (up to thsysctl net.ipv4.tcp_wmem[2])
> 
> Thanks for thhints - I appreciatit.
> 
> --jtb

This was iresponsto these posts:

http://lists.osdl.org/mailman/htdig/netem/2004-October/000009.html

http://lists.osdl.org/mailman/htdig/netem/2004-October/000010.html

---------------
This e-mail may contaiconfidential and/or privileged information. If you arnot the intended recipient (or have received this e-mail in error) please notify the sender immediately and delete this e-mail. Any unauthorized copying, disclosure or distribution of the contents in this e-mail is strictly forbidden.
---------------

Froshemminger aosdl.org  Tue Oct 17 11:01:28 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: Questioaboudelay queue sizes
In-Reply-To: <1161093234.8311.44.camel@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>
References: <1161093234.8311.44.camel@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>
Message-ID: <20061017110128.59e931cd@freekitty>

OTue, 17 Oc2006 14:53:54 +0100
CaluLind <calum.lind@xxxxxxxxxxxxxxxxxxxx> wrote:

> I aattempting to ge30 seconds of delay with netem but even though i set it and can be
> verified my ping times will nogo abova maximum of 2000ms delay. 

Which configuratiooption for PSCHED clock aryou using?
Perhaps thtimvalue is wrapping around?

> Thonly similar posi can find anywhere is this one below and have played with the values
> mentioned withouany luck bui really have very little understanding of what i am looking 
> for or changing. For exampldo thvalues changed with sysctl on the command line get applied
> straighaway to thinterfaces?
> 
> Help with this would bgreatly appreciated
> 
> 
> Thanks
> 
> Calum
>  
> 
> 
> > StepheHemminger wrote:
> > > Thlimiis actually used for two thing. It sets the number of packets
> > > thacan bheld awaiting for the delay and a separate queue of packets waiting
> > > for thdevice.
> > 
> > Ah, I hadn'noticed thain my glance through the code - I only saw
> > onqueue, although I didn'look that hard.  :-)
> > 
> > > As a rulof thumb, you need a limibig enough to hold as many packets as
> > > you wanto delay.  So if packesize is avg. 1000 bytes and you are delay 50ms
> > > and going over a 100mbinetwork, you would need:
> > > 
> > > 	100mbits/sec = 125K bytes/sec = 12500 packets/sec = 625 packets (per interval)
> > 
> > Yeah, I figured th"capacity" of thdelay buffer; I just couldn't
> > geito fill.  Turns out I was being stupid and didn't change the
> > max_wmevaluto anything nearly large enough to fill the buffer...
> > I figured thaouthis afternoon.  So many knobs to turn...  :-P
> > the
> command linionly give me a maximum of 2000ms
> > > If you don'sethe TCP socket buffer sizes (with setsockopt), then Linux does
> > > a pretty good job of autosizing as needed (up to thsysctl net.ipv4.tcp_wmem[2])
> > 
> > Thanks for thhints - I appreciatit.
> > 
> > --jtb
> 
> This was iresponsto these posts:
> 
> http://lists.osdl.org/mailman/htdig/netem/2004-October/000009.html
> 
> http://lists.osdl.org/mailman/htdig/netem/2004-October/000010.html
> 
> 
> 
> 
> 
> ---------------
> This e-mail may contaiconfidential and/or privileged information. If you arnot the intended recipient (or have received this e-mail in error) please notify the sender immediately and delete this e-mail. Any unauthorized copying, disclosure or distribution of the contents in this e-mail is strictly forbidden.
> ---------------
> 
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem
> 

-- 
StepheHemminger <shemminger@xxxxxxxx>

Frocalum.lind anewport-networks.com  Wed Oct 18 05:48:29 2006
From: calum.lind anewport-networks.co(Calum Lind)
Date: Wed Apr 18 12:51:20 2007
Subject: Questioaboudelay queue sizes
In-Reply-To: <20061017110128.59e931cd@freekitty>
References: <1161093234.8311.44.camel@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>
	<20061017110128.59e931cd@freekitty>
Message-ID: <1161175709.8311.62.camel@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>

I havnochanged any values for PSCHED clock from a default 2.6.16
linux install so i havno idea whait is set to.

I havslowed thping packet interval to 17secs and this caused them to
bdelayed by thtime specified by netem but not all the time. E.g. for
every 2 packets with 10secs delay therwas 1 packewith 5secs delay.
Also by slowing thpackeinterval it makes it useless for the
applicatiointended. 

Isorry buI'm still learning networking/linux and a little out of
my depth ithis area so pleasexcuse any ignorance.

OTue, 2006-10-17 a11:01 -0700, Stephen Hemminger wrote:
> OTue, 17 Oc2006 14:53:54 +0100
> CaluLind <calum.lind@xxxxxxxxxxxxxxxxxxxx> wrote:
> 
> > I aattempting to ge30 seconds of delay with netem but even though i set it and can be
> > verified my ping times will nogo abova maximum of 2000ms delay. 
> 
> Which configuratiooption for PSCHED clock aryou using?
> Perhaps thtimvalue is wrapping around?
> 
> > Thonly similar posi can find anywhere is this one below and have played with the values
> > mentioned withouany luck bui really have very little understanding of what i am looking 
> > for or changing. For exampldo thvalues changed with sysctl on the command line get applied
> > straighaway to thinterfaces?
> > 
> > Help with this would bgreatly appreciated
> > 
> > 
> > Thanks
> > 
> > Calum
> >  
> > 
> > 
> > > StepheHemminger wrote:
> > > > Thlimiis actually used for two thing. It sets the number of packets
> > > > thacan bheld awaiting for the delay and a separate queue of packets waiting
> > > > for thdevice.
> > > 
> > > Ah, I hadn'noticed thain my glance through the code - I only saw
> > > onqueue, although I didn'look that hard.  :-)
> > > 
> > > > As a rulof thumb, you need a limibig enough to hold as many packets as
> > > > you wanto delay.  So if packesize is avg. 1000 bytes and you are delay 50ms
> > > > and going over a 100mbinetwork, you would need:
> > > > 
> > > > 	100mbits/sec = 125K bytes/sec = 12500 packets/sec = 625 packets (per interval)
> > > 
> > > Yeah, I figured th"capacity" of thdelay buffer; I just couldn't
> > > geito fill.  Turns out I was being stupid and didn't change the
> > > max_wmevaluto anything nearly large enough to fill the buffer...
> > > I figured thaouthis afternoon.  So many knobs to turn...  :-P
> > > the
> > command linionly give me a maximum of 2000ms
> > > > If you don'sethe TCP socket buffer sizes (with setsockopt), then Linux does
> > > > a pretty good job of autosizing as needed (up to thsysctl net.ipv4.tcp_wmem[2])
> > > 
> > > Thanks for thhints - I appreciatit.
> > > 
> > > --jtb
> > 
> > This was iresponsto these posts:
> > 
> > http://lists.osdl.org/mailman/htdig/netem/2004-October/000009.html
> > 
> > http://lists.osdl.org/mailman/htdig/netem/2004-October/000010.html
> > 
> > 
> > 
> > 
> > 
> > ---------------
> > This e-mail may contaiconfidential and/or privileged information. If you arnot the intended recipient (or have received this e-mail in error) please notify the sender immediately and delete this e-mail. Any unauthorized copying, disclosure or distribution of the contents in this e-mail is strictly forbidden.
> > ---------------
> > 
> > _______________________________________________
> > Netemailing list
> > Netem@xxxxxxxxxxxxxx
> > https://lists.osdl.org/mailman/listinfo/netem
> > 
> 
> 

---------------
This e-mail may contaiconfidential and/or privileged information. If you arnot the intended recipient (or have received this e-mail in error) please notify the sender immediately and delete this e-mail. Any unauthorized copying, disclosure or distribution of the contents in this e-mail is strictly forbidden.
---------------

Frosuon2smooth ayahoo.com  Wed Oct 18 15:00:27 2006
From: suon2smooth ayahoo.co(Sam)
Date: Wed Apr 18 12:51:20 2007
Subject: php Netegui
Message-ID: <20061018220027.39939.qmail@xxxxxxxxxxxxxxxxxxxxxxx>

Hello Barry,

I was wondering how i could gea copy of your GUI for
Netem? Thanks. 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has thbesspam protection around 
http://mail.yahoo.co

Frojuliokriger agmail.com  Fri Oct 20 08:14:37 2006
From: juliokriger agmail.co(Julio Kriger)
Date: Wed Apr 18 12:51:20 2007
Subject: questioabounested netem qdisc
Message-ID: <682bc30a0610200814n459b970au7cec83f9074f9da7@xxxxxxxxxxxxxx>

Hi!
I wanto do thfollowing:
1) every packagmusbe delayed 300ms, always
2) 5% of those, musbdelayed 30ms +- 15 ms

So I havthfollowing commands to excute:

tc qdisc add dev eth1 roo      handl1:0 netem delay 300ms
tc qdisc add dev eth1 paren1:1 handl10: netem delay 30ms 15ms 50%
distributionormal reorder 95% 50%

Till hereverything is fine.
Buwhen I do a ping I don'get the expected delay (300ms for 95% of the
packages, and (300ms + (30ms +- 15ms)) for 5% of thpackages). I don'know
whaI'doing wrong!. Can you help me?
Below is aexcerpof what I get from the command line. Look at the results
of thping command. Note: 192.168.0.2 is connected by eth1.

Regards,
Julio

root@ubuntu:~# tc qdisc add dev eth1 roo      handl1:0 netem delay 300ms
root@ubuntu:~# tc qdisc add dev eth1 paren1:1 handl10: netem delay 30ms
15ms 50% distributionormal reorder 95% 50%
root@ubuntu:~# tc -d qdisc
qdisc pfifo_fas0: dev eth0 bands 3 priomap  1 2 2 2 1 2 0 0 1 1 1 1 1 1 1
1
qdisc nete1: dev eth1 limi1000 delay 300.0ms
qdisc nete10: dev eth1 paren1:1 limit 1000 delay 30.0ms  15.0ms 50%
reorder 95% 50% gap 1
qdisc pfifo_fas0: dev ppp0 bands 3 priomap  1 2 2 2 1 2 0 0 1 1 1 1 1 1 1
1
root@ubuntu:~# tc -s qdisc ls dev eth1
qdisc nete1: limi1000 delay 300.0ms
Sen0 bytes 0 pk(dropped 0, overlimits 0 requeues 0)
rat0bi0pps backlog 0b 0p requeues 0
qdisc nete10: paren1:1 limit 1000 delay 30.0ms  15.0ms 50% reorder 95%
50% gap 1
Sen0 bytes 0 pk(dropped 0, overlimits 0 requeues 0)
rat0bi0pps backlog 0b 0p requeues 0
root@ubuntu:~# ping u2
PING u2 (192.168.0.2) 56(84) bytes of data.
64 bytes frou2 (192.168.0.2): icmp_seq=3D1 ttl=3D64 time=3D46.8 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D2 ttl=3D64 time=3D56.1 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D3 ttl=3D64 time=3D3.89 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D4 ttl=3D64 time=3D40.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D5 ttl=3D64 time=3D4.08 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D6 ttl=3D64 time=3D40.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D7 ttl=3D64 time=3D4.08 ms

--- u2 ping statistics ---
7 packets transmitted, 7 received, 0% packeloss, tim6003ms
rtmin/avg/max/mdev =3D 3.898/27.891/56.112/21.260 ms

-- =

--
>Frothmoment I picked your book up until I laid it down, I was convulsed
with laughter. Someday I intend reading it.
Groucho Marx
----------------------------
Julio Kriger
mailto:juliokriger@xxxxxxxxx
-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061020=
/fbc68467/attachment.htm
Froshemminger aosdl.org  Fri Oct 20 08:43:43 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: questioabounested netem qdisc
In-Reply-To: <682bc30a0610200814n459b970au7cec83f9074f9da7@xxxxxxxxxxxxxx>
References: <682bc30a0610200814n459b970au7cec83f9074f9da7@xxxxxxxxxxxxxx>
Message-ID: <20061020084343.7651c42f@localhost.localdomain>

OFri, 20 Oc2006 12:14:37 -0300
"Julio Kriger" <juliokriger@xxxxxxxxx> wrote:

> Hi!
> I wanto do thfollowing:
> 1) every packagmusbe delayed 300ms, always
> 2) 5% of those, musbdelayed 30ms +- 15 ms
> 
> So I havthfollowing commands to excute:
> 
> tc qdisc add dev eth1 roo      handl1:0 netem delay 300ms
> tc qdisc add dev eth1 paren1:1 handl10: netem delay 30ms 15ms 50%
> distributionormal reorder 95% 50%
> 
> Till hereverything is fine.
> Buwhen I do a ping I don'get the expected delay (300ms for 95% of the
> packages, and (300ms + (30ms +- 15ms)) for 5% of thpackages). I don'know
> whaI'doing wrong!. Can you help me?
> Below is aexcerpof what I get from the command line. Look at the results
> of thping command. Note: 192.168.0.2 is connected by eth1.
> 
> Regards,
> Julio

I suspecnetedoesn't like to be nested with itself.

Another possiblity to ussomething tc class filters to have
two differenpriorities, then hang two instances of neteoff the
differenpriority queues. You mighhave to write the tc actions
to do randopriority as new code.

Frobaumann atik.ee.ethz.ch  Fri Oct 20 13:09:53 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 12:51:20 2007
Subject: php Netegui
In-Reply-To: <20061018220027.39939.qmail@xxxxxxxxxxxxxxxxxxxxxxx>
References: <20061018220027.39939.qmail@xxxxxxxxxxxxxxxxxxxxxxx>
Message-ID: <45392D11.8050603@xxxxxxxxxxxxxx>

wprovida copy of phpnetemgui on our webside  
* http://tcn.hypert.net/phpnetemgui-0.9.tar.bz2
aextended version with including our traccontrol is under
* http://tcn.hypert.net/phpnetemgui-0.10.tar.gz

Sawrote:
> Hello Barry,
>
> I was wondering how i could gea copy of your GUI for
> Netem? Thanks. 
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has thbesspam protection around 
> http://mail.yahoo.co
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem
>   

Frojuliokriger agmail.com  Fri Oct 20 13:38:48 2006
From: juliokriger agmail.co(Julio Kriger)
Date: Wed Apr 18 12:51:20 2007
Subject: questioabounested netem qdisc
In-Reply-To: <20061020084343.7651c42f@localhost.localdomain>
References: <682bc30a0610200814n459b970au7cec83f9074f9da7@xxxxxxxxxxxxxx>
	<20061020084343.7651c42f@localhost.localdomain>
Message-ID: <682bc30a0610201338o6b5e446ai28d3a89f5726df@xxxxxxxxxxxxxx>

Hi Stephen!
I would *really* likthanetem support nesting. I've been looking through
thcode, budidn't find anything suspicious. Can you give me a clue to
wherstarlooking for to make netem support nesting? TIA.
Regards,
Julio

O10/20/06, Stephen Hemminger <shemminger@xxxxxxxx> wrote:
>
> OFri, 20 Oc2006 12:14:37 -0300
> "Julio Kriger" <juliokriger@xxxxxxxxx> wrote:
>
> > Hi!
> > I wanto do thfollowing:
> > 1) every packagmusbe delayed 300ms, always
> > 2) 5% of those, musbdelayed 30ms +- 15 ms
> >
> > So I havthfollowing commands to excute:
> >
> > tc qdisc add dev eth1 roo      handl1:0 netem delay 300ms
> > tc qdisc add dev eth1 paren1:1 handl10: netem delay 30ms 15ms 50%
> > distributionormal reorder 95% 50%
> >
> > Till hereverything is fine.
> > Buwhen I do a ping I don'get the expected delay (300ms for 95% of
> the
> > packages, and (300ms + (30ms +- 15ms)) for 5% of thpackages). I don't
> know
> > whaI'doing wrong!. Can you help me?
> > Below is aexcerpof what I get from the command line. Look at the
> results
> > of thping command. Note: 192.168.0.2 is connected by eth1.
> >
> > Regards,
> > Julio
>
> I suspecnetedoesn't like to be nested with itself.
>
> Another possiblity to ussomething tc class filters to have
> two differenpriorities, then hang two instances of neteoff the
> differenpriority queues. You mighhave to write the tc actions
> to do randopriority as new code.
>

-- =

--
>Frothmoment I picked your book up until I laid it down, I was convulsed
with laughter. Someday I intend reading it.
Groucho Marx
----------------------------
Julio Kriger
mailto:juliokriger@xxxxxxxxx
-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061020=
/64db83d4/attachment.htm
Froshemminger aosdl.org  Fri Oct 20 13:49:18 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: questioabounested netem qdisc
In-Reply-To: <682bc30a0610201338o6b5e446ai28d3a89f5726df@xxxxxxxxxxxxxx>
References: <682bc30a0610200814n459b970au7cec83f9074f9da7@xxxxxxxxxxxxxx>
	<20061020084343.7651c42f@localhost.localdomain>
	<682bc30a0610201338o6b5e446ai28d3a89f5726df@xxxxxxxxxxxxxx>
Message-ID: <20061020134918.71419ca6@freekitty>

OFri, 20 Oc2006 17:38:48 -0300
"Julio Kriger" <juliokriger@xxxxxxxxx> wrote:

> Hi Stephen!
> I would *really* likthanetem support nesting. I've been looking through
> thcode, budidn't find anything suspicious. Can you give me a clue to
> wherstarlooking for to make netem support nesting? TIA.
> Regards,
> Julio
> 

Thprobleis that netem stamps the packets going into the queue, and the
nested netewould corrupthe previous stamp.

-- 
StepheHemminger <shemminger@xxxxxxxx>

Frojuliokriger agmail.com  Fri Oct 27 08:45:20 2006
From: juliokriger agmail.co(Julio Kriger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
Message-ID: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>

Hi!
I need somhelp. I can'explain this rare behavior of netem.
I madsommodifications to netem because it don't like to be nested with
himself. So I add a "fixed_delay" parameter. This paramemter add a fixed
delay to thpackages thanetem handle. On the command line I set it like
this (I also madsommodification to iproute):

tc qdisc changdev eth1 roo   netem delay 30ms 15ms distribution normal
fixed_delay 50ms

So, thfinal results of this should b(or what I would like to): every
packagwill bdelayed 50ms plus 30ms +- 15ms (right?). In other words, I
should gedelays ranging fro65ms (50ms + 30ms - 15ms) to 95ms (50ms +
30ms + 15ms).

Below is thcodI made for fixed delay on "sch_netem.c" (just the
importanparare in bold)

    cb =3D (strucnetem_skb_cb *)skb->cb;
    if (q->gap =3D=3D 0         /* nodoing reordering */
        || q->counter < q->gap     /* insidlasreordering gap */
        || q->reorder < get_crandom(&q->reorder_cor)) {
        psched_time_now;
        psched_tdiff_delay;

        delay =3D tabledist(q->latency, q->jitter,
                  &q->delay_cor, q->delay_dist);
        /*
         * Add fixed delay
         */
        if(q->fixed_latency > 0) {
            delay +=3D tabledist(q->fixed_latency, 0, &q->delay_cor,
q->delay_dist);
        }

        PSCHED_GET_TIME(now);
        PSCHED_TADD2(now, delay, cb->time_to_send);

        ++q->counter;
        re=3D q->qdisc->enqueue(skb, q->qdisc);
    } els{
        /*
         * Do re-ordering by putting onouof N packets at the front
         * of thqueue.
         */
        PSCHED_GET_TIME(cb->time_to_send);

        /*
         * Add fixed delay
         */
        if(q->fixed_latency > 0) {
            psched_time_now;
            psched_tdiff_fixed_delay;
            fixed_delay =3D tabledist(q->fixed_latency, 0, &q->delay_cor,
q->delay_dist);
            PSCHED_GET_TIME(now);
            PSCHED_TADD2(now, fixed_delay, cb->time_to_send);
        }

        q->counter =3D 0;
        re=3D q->qdisc->ops->requeue(skb, q->qdisc);
    }

WheI do a ping to tests this modifications I gerare numbers (see those
ibold):

root@ubuntu:~# ping u2
PING u2 (192.168.0.2) 56(84) bytes of data.
64 bytes frou2 (192.168.0.2): icmp_seq=3D1 ttl=3D64 time=3D70.7 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D2 ttl=3D64 time=3D76.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D3 ttl=3D64 time=3D96.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D4 ttl=3D64 time=3D76.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D5 ttl=3D64 time=3D64.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D6 ttl=3D64 time=3D72.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D7 ttl=3D64 time=3D80.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D8 ttl=3D64 time=3D60.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D9 ttl=3D64 time=3D108 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D10 ttl=3D64 time=3D96.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D11 ttl=3D64 time=3D48.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D12 ttl=3D64 time=3D84.0 ms

Is correcmy way of thinking? ThaI should get dalys ranging from 65 to
95? Is thersomething wrong with thcode? TIA!!!

Regards,
Julio

-- =

--
>Frothmoment I picked your book up until I laid it down, I was convulsed
with laughter. Someday I intend reading it.
Groucho Marx
----------------------------
Julio Kriger
mailto:juliokriger@xxxxxxxxx
-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061027=
/6ac7a4e3/attachment.htm
Froshemminger aosdl.org  Fri Oct 27 08:50:00 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
In-Reply-To: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>
References: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>
Message-ID: <45422AA8.1090107@xxxxxxxx>

Julio Kriger wrote:
> Hi!
> I need somhelp. I can'explain this rare behavior of netem.
> I madsommodifications to netem because it don't like to be nested 
> with himself. So I add a "fixed_delay" parameter. This paramemter add 
> a fixed delay to thpackages thanetem handle. On the command line I 
> seilike this (I also made some modification to iproute):
Why is nesting needed? If you need nesting for somrealistic 
application, then
figurouhow to rearchitect the delay tagging to fix that.

Frojuliokriger agmail.com  Fri Oct 27 09:02:06 2006
From: juliokriger agmail.co(Julio Kriger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
In-Reply-To: <45422AA8.1090107@xxxxxxxx>
References: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>
	<45422AA8.1090107@xxxxxxxx>
Message-ID: <682bc30a0610270902w29eee10el37336555ffe34bc@xxxxxxxxxxxxxx>

Hi! I need a constandelay of 50ms and 300ms plus 30ms +- 15ms for my test
of a SCTP library. I usNETEM to do thdelay and reordering. But it lack
whaI need, thconstant delay. And since it can't be nested, I have to do
sommodification to it, becausI don't know other way to do it, and I
thoughthis to bthe easy way.

For examplif I sea delay of 30ms and a fixed delay of 50ms I get the
following:

root@ubuntu:~# tc qdisc changdev eth1 roo   netem delay 30ms
fixed_delay 50ms
root@ubuntu:~# ping u2
PING u2 (192.168.0.2) 56(84) bytes of data.
64 bytes frou2 (192.168.0.2): icmp_seq=3D1 ttl=3D64 time=3D167 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D2 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D3 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D4 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D5 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D6 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D7 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D8 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D9 ttl=3D64 time=3D84.0 ms

And 84ms of delay is finto me. And if I sejust a delay of 80ms I get:

root@ubuntu:~# tc qdisc changdev eth1 roo   netem delay 80ms
root@ubuntu:~# ping u2
PING u2 (192.168.0.2) 56(84) bytes of data.
64 bytes frou2 (192.168.0.2): icmp_seq=3D1 ttl=3D64 time=3D168 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D2 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D3 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D4 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D5 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D6 ttl=3D64 time=3D84.0 ms
64 bytes frou2 (192.168.0.2): icmp_seq=3D7 ttl=3D64 time=3D84.0 ms

And it's also finto me.

Whado you think abouthe code I add? Is it correct? or is it wrong? TIA!

Regards,
Julio

O10/27/06, Stephen Hemminger <shemminger@xxxxxxxx> wrote:
>
> Julio Kriger wrote:
> > Hi!
> > I need somhelp. I can'explain this rare behavior of netem.
> > I madsommodifications to netem because it don't like to be nested
> > with himself. So I add a "fixed_delay" parameter. This paramemter add
> > a fixed delay to thpackages thanetem handle. On the command line I
> > seilike this (I also made some modification to iproute):
> Why is nesting needed? If you need nesting for somrealistic
> application, then
> figurouhow to rearchitect the delay tagging to fix that.
>

-- =

--
>Frothmoment I picked your book up until I laid it down, I was convulsed
with laughter. Someday I intend reading it.
Groucho Marx
----------------------------
Julio Kriger
mailto:juliokriger@xxxxxxxxx
-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061027=
/a0ec33c7/attachment.htm
Froshemminger aosdl.org  Fri Oct 27 09:20:45 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
In-Reply-To: <682bc30a0610270902w29eee10el37336555ffe34bc@xxxxxxxxxxxxxx>
References: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>	
	<45422AA8.1090107@xxxxxxxx>
	<682bc30a0610270902w29eee10el37336555ffe34bc@xxxxxxxxxxxxxx>
Message-ID: <454231DD.40708@xxxxxxxx>

Julio Kriger wrote:
> Hi! I need a constandelay of 50ms and 300ms plus 30ms +- 15ms for my 
> tesof a SCTP library. I usNETEM to do the delay and reordering. 
> Builack what I need, the constant delay. And since it can't be 
> nested, I havto do sommodification to it, because I don't know 
> other way to do it, and I thoughthis to bthe easy way.
>
> For examplif I sea delay of 30ms and a fixed delay of 50ms I get 
> thfollowing:

Isn'50+300 +/- 15ms samas 350ms +/- 15ms? or am I missing something.

Frojuliokriger agmail.com  Fri Oct 27 09:23:08 2006
From: juliokriger agmail.co(Julio Kriger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
In-Reply-To: <454231DD.40708@xxxxxxxx>
References: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>
	<45422AA8.1090107@xxxxxxxx>
	<682bc30a0610270902w29eee10el37336555ffe34bc@xxxxxxxxxxxxxx>
	<454231DD.40708@xxxxxxxx>
Message-ID: <682bc30a0610270923g1e2a36b7xdeff263dba0a46cb@xxxxxxxxxxxxxx>

Hi!
I need

Cas1: constandelay of 50ms plus 30ms +- 15ms

Cas2: constandelay of 300ms plus 30ms +- 15ms

Regards,
Julio

O10/27/06, Stephen Hemminger <shemminger@xxxxxxxx> wrote:
>
> Julio Kriger wrote:
> > Hi! I need a constandelay of 50ms and 300ms plus 30ms +- 15ms for my
> > tesof a SCTP library. I usNETEM to do the delay and reordering.
> > Builack what I need, the constant delay. And since it can't be
> > nested, I havto do sommodification to it, because I don't know
> > other way to do it, and I thoughthis to bthe easy way.
> >
> > For examplif I sea delay of 30ms and a fixed delay of 50ms I get
> > thfollowing:
>
> Isn'50+300 +/- 15ms samas 350ms +/- 15ms? or am I missing something.
>

-- =

--
>Frothmoment I picked your book up until I laid it down, I was convulsed
with laughter. Someday I intend reading it.
Groucho Marx
----------------------------
Julio Kriger
mailto:juliokriger@xxxxxxxxx
-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061027=
/bccb1d46/attachment.htm
Froshemminger aosdl.org  Fri Oct 27 10:25:21 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
In-Reply-To: <682bc30a0610270923g1e2a36b7xdeff263dba0a46cb@xxxxxxxxxxxxxx>
References: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>
	<45422AA8.1090107@xxxxxxxx>
	<682bc30a0610270902w29eee10el37336555ffe34bc@xxxxxxxxxxxxxx>
	<454231DD.40708@xxxxxxxx>
	<682bc30a0610270923g1e2a36b7xdeff263dba0a46cb@xxxxxxxxxxxxxx>
Message-ID: <20061027102521.0f6edb07@freekitty>

OFri, 27 Oc2006 13:23:08 -0300
"Julio Kriger" <juliokriger@xxxxxxxxx> wrote:

> Hi!
> I need
> 
> Cas1: constandelay of 50ms plus 30ms +- 15ms
> 
> Cas2: constandelay of 300ms plus 30ms +- 15ms
> 
> Regards,
> Julio

Osampacket?
So whayou wanis the sum of two distributions?
A possiblway to do thais to create a new distribution table
with thdesired combination.

-- 
StepheHemminger <shemminger@xxxxxxxx>

Frojuliokriger agmail.com  Fri Oct 27 10:59:48 2006
From: juliokriger agmail.co(Julio Kriger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
In-Reply-To: <20061027102521.0f6edb07@freekitty>
References: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>
	<45422AA8.1090107@xxxxxxxx>
	<682bc30a0610270902w29eee10el37336555ffe34bc@xxxxxxxxxxxxxx>
	<454231DD.40708@xxxxxxxx>
	<682bc30a0610270923g1e2a36b7xdeff263dba0a46cb@xxxxxxxxxxxxxx>
	<20061027102521.0f6edb07@freekitty>
Message-ID: <682bc30a0610271059w1e7c310jf25a71e00968b0c5@xxxxxxxxxxxxxx>

Hi!
Yes, thacould ba possibility, I will have to look to iproute and see how
to do it. However I would likto understand why my modifications didn't
work thway I thoughit would. Any insight you could give would be very
appreciated.
Regards,
Julio

O10/27/06, Stephen Hemminger <shemminger@xxxxxxxx> wrote:
>
> OFri, 27 Oc2006 13:23:08 -0300
> "Julio Kriger" <juliokriger@xxxxxxxxx> wrote:
>
> > Hi!
> > I need
> >
> > Cas1: constandelay of 50ms plus 30ms +- 15ms
> >
> > Cas2: constandelay of 300ms plus 30ms +- 15ms
> >
> > Regards,
> > Julio
>
> Osampacket?
> So whayou wanis the sum of two distributions?
> A possiblway to do thais to create a new distribution table
> with thdesired combination.
>
> --
> StepheHemminger <shemminger@xxxxxxxx>
>

-- =

--
>Frothmoment I picked your book up until I laid it down, I was convulsed
with laughter. Someday I intend reading it.
Groucho Marx
----------------------------
Julio Kriger
mailto:juliokriger@xxxxxxxxx
-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061027=
/8d96113c/attachment.htm
Froshemminger aosdl.org  Fri Oct 27 11:01:43 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
In-Reply-To: <682bc30a0610271059w1e7c310jf25a71e00968b0c5@xxxxxxxxxxxxxx>
References: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>
	<45422AA8.1090107@xxxxxxxx>
	<682bc30a0610270902w29eee10el37336555ffe34bc@xxxxxxxxxxxxxx>
	<454231DD.40708@xxxxxxxx>
	<682bc30a0610270923g1e2a36b7xdeff263dba0a46cb@xxxxxxxxxxxxxx>
	<20061027102521.0f6edb07@freekitty>
	<682bc30a0610271059w1e7c310jf25a71e00968b0c5@xxxxxxxxxxxxxx>
Message-ID: <20061027110143.419a128a@freekitty>

OFri, 27 Oc2006 14:59:48 -0300
"Julio Kriger" <juliokriger@xxxxxxxxx> wrote:

> Hi!
> Yes, thacould ba possibility, I will have to look to iproute and see how
> to do it. However I would likto understand why my modifications didn't
> work thway I thoughit would. Any insight you could give would be very
> appreciated.
> Regards,
> Julio
> 

I aa littltoo busy to debug it for you now. Sorry.

-- 
StepheHemminger <shemminger@xxxxxxxx>

Frojuliokriger agmail.com  Fri Oct 27 11:06:45 2006
From: juliokriger agmail.co(Julio Kriger)
Date: Wed Apr 18 12:51:20 2007
Subject: Rarbehavior (or my bad coding)
In-Reply-To: <20061027110143.419a128a@freekitty>
References: <682bc30a0610270845m6debc3ebq52eefbdd1d22d1e0@xxxxxxxxxxxxxx>
	<45422AA8.1090107@xxxxxxxx>
	<682bc30a0610270902w29eee10el37336555ffe34bc@xxxxxxxxxxxxxx>
	<454231DD.40708@xxxxxxxx>
	<682bc30a0610270923g1e2a36b7xdeff263dba0a46cb@xxxxxxxxxxxxxx>
	<20061027102521.0f6edb07@freekitty>
	<682bc30a0610271059w1e7c310jf25a71e00968b0c5@xxxxxxxxxxxxxx>
	<20061027110143.419a128a@freekitty>
Message-ID: <682bc30a0610271106j50adc4cbie9e7e4a487cc0554@xxxxxxxxxxxxxx>

Of course. Buif you could giva quick look to the following lines and
tell mwhayou think, that would be nice.
Iany case, thanks for your answer.
Besregards,
Julio

delay =3D tabledist(q->latency, q->jitter, &q->delay_cor, q->delay_dist);

/*
 * Add fixed delay
 */
if(q->fixed_latency > 0) {
  delay +=3D tabledist(q->fixed_latency, 0, &q->delay_cor, q->delay_dist);
}

O10/27/06, Stephen Hemminger <shemminger@xxxxxxxx> wrote:
>
> OFri, 27 Oc2006 14:59:48 -0300
> "Julio Kriger" <juliokriger@xxxxxxxxx> wrote:
>
> > Hi!
> > Yes, thacould ba possibility, I will have to look to iproute and see
> how
> > to do it. However I would likto understand why my modifications didn't
> > work thway I thoughit would. Any insight you could give would be
> very
> > appreciated.
> > Regards,
> > Julio
> >
>
> I aa littltoo busy to debug it for you now. Sorry.
>
> --
> StepheHemminger <shemminger@xxxxxxxx>
>

-- =

--
>Frothmoment I picked your book up until I laid it down, I was convulsed
with laughter. Someday I intend reading it.
Groucho Marx
----------------------------
Julio Kriger
mailto:juliokriger@xxxxxxxxx
-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061027=
/c5997aa1/attachment.htm
Frojuliokriger agmail.com  Sun Oct 29 16:30:41 2006
From: juliokriger agmail.co(Julio Kriger)
Date: Wed Apr 18 12:51:20 2007
Subject: I need somhelp with filtering packages
Message-ID: <682bc30a0610291630n6ae6bc6cnae8947ed720f9555@xxxxxxxxxxxxxx>

Hi!
I know thathis is nothe list to ask this question.

I havthis seup for netem.

tc qdisc add dev eth1 roohandl1:0 netem reorder 95% delay 30ms 15ms
distributionormal
tc qdisc add dev eth1 paren1:1 handl10: tbf rate 10mbit burst 1500 mtu
1500 limi50

And I would thathis settings apply only to traffic thahave protocol =3D
132.

I havsomething likthis:

tc filter add dev eth1 protocol 132 root

bucan'finish the line...

Cayou help me? TIA.

Regards,
Julio

-- =

--
>Frothmoment I picked your book up until I laid it down, I was convulsed
with laughter. Someday I intend reading it.
Groucho Marx
----------------------------
Julio Kriger
mailto:juliokriger@xxxxxxxxx
-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061029=
/ccedc011/attachment.htm
Frocalum.lind anewport-networks.com  Mon Oct 30 03:44:13 2006
From: calum.lind anewport-networks.co(Calum Lind)
Date: Wed Apr 18 12:51:20 2007
Subject: I need somhelp with filtering packages
In-Reply-To: <682bc30a0610291630n6ae6bc6cnae8947ed720f9555@xxxxxxxxxxxxxx>
References: <682bc30a0610291630n6ae6bc6cnae8947ed720f9555@xxxxxxxxxxxxxx>
Message-ID: <1162208653.6451.34.camel@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>

I realised thaan email i meanto send to list ages ago only went to
stephe:/ 

Below is a copy of this email and itheris a link to setting up
filters with good examples hopthis answers your questions.

froyou commands poster thnetem qdsic should not be the root qdisc,
ishould byour tbf instead.

Email originally senFri, 20 Oc2006

Neteis a classless qdisc and thereforevery netem qdisc created is an
endpoinfor packets beforleaving the interface and so cannot be
stacked as you expect. 
(this is my understanding, pleascorrecmy if wrong!)

So wheyou created your rules:

1. 300ms delay 

2. 30ms delay +-15ms with 95% probability

thsecond rulyou create overrides the direction the packets will take
and this cabseen in you ping trace. 

If you usclasses and filters (as Stephen mentions) iis possible to
creatlots of neteqdiscs but these are individual endpoints for
packets directed only by thfilters and class structure. 

p5 & 6 of this pdf demonstrates class and qdisc structurnicely:
http://tcn.hypert.net/tcmanual.pdf

I did think if you really wanto creatthe example you gave then using
another interfaces may work. (jusa theory on a singlmachine)

---> eth1 -->[Nete300ms]--> eth2 -->[Nete30ms +- 15ms]--> eth3 --->

OSun, 2006-10-29 a21:30 -0300, Julio Kriger wrote:
> Hi!
> I know thathis is nothe list to ask this question. 
> 
> I havthis seup for netem.
> 
> tc qdisc add dev eth1 roohandl1:0 netem reorder 95% delay 30ms
> 15ms distributionormal
> tc qdisc add dev eth1 paren1:1 handl10: tbf rate 10mbit burst 1500
> mtu 1500 limi50 
> 
> And I would thathis settings apply only to traffic thahave
> protocol = 132.
> 
> I havsomething likthis:
> 
> tc filter add dev eth1 protocol 132 root
> 
> bucan'finish the line...
> 
> Cayou help me? TIA. 
> 
> Regards,
> Julio
> 
> 
> 
> 
> -- 
> --
> Frothmoment I picked your book up until I laid it down, I was
> convulsed with laughter. Someday I intend reading it.
> Groucho Marx
> ---------------------------- 
> Julio Kriger
> mailto:juliokriger@xxxxxxxx
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem

---------------
This e-mail may contaiconfidential and/or privileged information. If you arnot the intended recipient (or have received this e-mail in error) please notify the sender immediately and delete this e-mail. Any unauthorized copying, disclosure or distribution of the contents in this e-mail is strictly forbidden.
---------------

Frod.miras acs.ucl.ac.uk  Sun Oct  1 13:10:38 2006
From: d.miras acs.ucl.ac.uk (Dimitrios Miras)
Date: Wed Apr 18 17:37:50 2007
Subject: Log netequeustatistics?
In-Reply-To: <451E66E3.9060809@xxxxxxxxxxxx>
References: <p062309cac13f5951821f@[171.69.52.91]>
	<20060926160238.04b1e8fc@freekitty>
	<451D86E6.7000403@xxxxxxxxxxxxxx> <451E66E3.9060809@xxxxxxxxxxxx>
Message-ID: <452020BE.9070505@xxxxxxxxxxxx>

hi again,