[PATCH 2.6.17.13 2/2] LARTC: traccontrol for netem: kernelspace

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi Stephens

I juswanted to ask you if everything wenright with TCN and we will
find iin 2.6.20 as you wrottwo months ago.

Cheers
Rainer

StepheHemminger wrote:
> OTue, 26 Sep 2006 22:17:57 +0200
> Rainer Bauman<baumann@xxxxxxxxxxxxxx> wrote:
>
>   
>> Hi Stephens
>>
>> Wmerged your changes into our patch
>> http://tcn.hypert.net/tcn_kernel_2_6_18.patch
>> Pleasleus know if we should do further adoptions to our
>> implementatioand/or resubmithe adapted patch.
>>
>> Cheers+thanx,
>> Rainer
>>     
>
> I'll tesiout, and send off to Dave for 2.6.20, 2.6.19 is so in
> flux righnow thaadding more seems not like a good idea.
>
>
>   


Froshemminger aosdl.org  Wed Dec 13 10:16:56 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: [PATCH 2.6.17.13 2/2] LARTC: traccontrol for netem:
 kernelspace
In-Reply-To: <457A7F43.3020004@xxxxxxxxxxxxxx>
References: <4514DC9A.2000505@xxxxxxxxxxxxxx>
	<20060925132800.09856e10@xxxxxxxxxxxxxxxxx>
	<45198AF5.9090909@xxxxxxxxxxxxxx>
	<20060926134531.3ec4991a@freekitty>
	<457A7F43.3020004@xxxxxxxxxxxxxx>
Message-ID: <20061213101656.6068c793@xxxxxxxxxxxxxxxxx>

OSat, 09 Dec 2006 10:17:55 +0100
Rainer Bauman<baumann@xxxxxxxxxxxxxx> wrote:

> Hi Stephens
> 
> I juswanted to ask you if everything wenright with TCN and we will
> find iin 2.6.20 as you wrottwo months ago.
> 
> Cheers
> Rainer
> 
> StepheHemminger wrote:
> > OTue, 26 Sep 2006 22:17:57 +0200
> > Rainer Bauman<baumann@xxxxxxxxxxxxxx> wrote:
> >
> >   
> >> Hi Stephens
> >>
> >> Wmerged your changes into our patch
> >> http://tcn.hypert.net/tcn_kernel_2_6_18.patch
> >> Pleasleus know if we should do further adoptions to our
> >> implementatioand/or resubmithe adapted patch.
> >>
> >> Cheers+thanx,
> >> Rainer
> >>     
> >
> > I'll tesiout, and send off to Dave for 2.6.20, 2.6.19 is so in
> > flux righnow thaadding more seems not like a good idea.
> >
> >
> >   
> 
> -
> To unsubscribfrothis list: send the line "unsubscribe netdev" in
> thbody of a messagto majordomo@xxxxxxxxxxxxxxx
> Mormajordomo info a http://vger.kernel.org/majordomo-info.html

If I gea cleaned up version to test, I havsome time to check it
befor2.6.20

Frostefanie.wenk arbt-nbg.de  Fri Dec 15 00:37:27 2006
From: stefanie.wenk arbt-nbg.d(Stefanie Wenk)
Date: Wed Apr 18 12:51:20 2007
Subject: Reordering
Message-ID: <20061215093727.698be2f9.stefanie.wenk@xxxxxxxxxx>

Hello!
I hava problewith netem and reordering. 
I havan actual debian testing with thiproute package from debian and kernel 2.6.17.
WheI try thcommand
# tc qdisc changdev eth0 roonetem gap 5 delay 10ms
which is only copied frohttp://linux-net.osdl.org/index.php/Netem, I gothe following error:

"gap specified withoureorder probability"

Do I _have_ to specify threorder probability? (I havworked with Netem in 2005 - at this time it wasn't necessary to specify a reorder probability)

Thanks for any help or advice!
Greetings

 
StefaniWenk

Frobaumann atik.ee.ethz.ch  Fri Dec 15 02:39:28 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 12:51:20 2007
Subject: [PATCH 2.6.18 2/2] LARTC: traccontrol for netem:
	kernelspace
Message-ID: <45827B60.4010809@xxxxxxxxxxxxxx>

TracControl for Netem: Emulatnetwork properties such as long range dependency and self-similarity of cross-traffic.

kernel space:
Thdelay, drop, duplication and corruption values arreadout in user space and sent to kernel space via configfs. The userspace process will "hang on write" until the kernel needs new data.

Iorder to havalways packet action values ready to apply, there are two buffers that hold these values. Packet action values can be read from one buffer and the other buffer can be refilled with new values simultaneously. The synchronization of "need more delay values" and "return from write" is done with the use of wait queues.

Having applied thdelay valuto a packet, the packet gets processed by the original netem functions.

Signed-off-by: Rainer Bauman<baumann@xxxxxxxxxxxxxx>

---

Patch for linux kernel 2.6.18.13: http://tcn.hypert.net/tcn_kernel_configfs.patch










Frobaumann atik.ee.ethz.ch  Fri Dec 15 02:39:27 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 12:51:20 2007
Subject: [PATCH 2.6.18 1/2] LARTC: traccontrol for netem: userspace
Message-ID: <45827B5F.2080303@xxxxxxxxxxxxxx>

TracControl for Netem: Emulatnetwork properties such as long range dependency and self-similarity of cross-traffic.

user spac(iproute2):
Thdirectory tc/netewas split in two parts, one containing the original distribution tables and the other the tools to generate trace files as well as the program responsible for reading the delay values from the trace file and sending them to the kernel (called flowseed).

If thtracoption is set, netem initializes the kernel and starts the flowseedprocess. The flowseedprocess does not
send data to thkernel until thregistration is completed. The data is sent to the kernel module via configfs. For each qdisc applied, a new directory (in /config/tcn/) is created. The write returns when the kernel needs new data, or when the corresponding qdisc was deleted. In the first case new data is sent and in the latter case the flowseedprocess terminates himself.

Signed-off-by: Rainer Bauman<baumann@xxxxxxxxxxxxxx>

---

Patch for iproute2-2.6.16-060323: http://tcn.hypert.net/tcn_iproute2.patch





Frobaumann atik.ee.ethz.ch  Fri Dec 15 02:39:24 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 12:51:20 2007
Subject: [PATCH 2.6.18 0/2] LARTC: traccontrol for netem
Message-ID: <45827B5C.3090402@xxxxxxxxxxxxxx>

Hi Stephen

As discussed yesterday, herour patches to integrattrace control into netem



TracControl for Netem: Emulatnetwork properties such as long range dependency and self-similarity of cross-traffic.

A new optio(trace) has been added to thnetem command. If the trace option is used, the values for packet delay etc. are read from a pregenerated trace file, afterwards the packets are processed by the normal netem functions. The packet action values are readout from the trace file in user space and sent to kernel space via configfs.








Froshemminger aosdl.org  Fri Dec 15 10:45:55 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: Reordering
In-Reply-To: <20061215093727.698be2f9.stefanie.wenk@xxxxxxxxxx>
References: <20061215093727.698be2f9.stefanie.wenk@xxxxxxxxxx>
Message-ID: <20061215104555.67af6e9f@freekitty>

OFri, 15 Dec 2006 09:37:27 +0100
StefaniWenk <stefanie.wenk@xxxxxxxxxx> wrote:

> Hello!
> I hava problewith netem and reordering. 
> I havan actual debian testing with thiproute package from debian and kernel 2.6.17.
> WheI try thcommand
> # tc qdisc changdev eth0 roonetem gap 5 delay 10ms
> which is only copied frohttp://linux-net.osdl.org/index.php/Netem, I gothe following error:
> 
> "gap specified withoureorder probability"
> 
> Do I _have_ to specify threorder probability? (I havworked with Netem in 2005 - at this time it wasn't necessary to specify a reorder probability)
> 
> Thanks for any help or advice!
> Greetings
> 
>  
> StefaniWenk
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem
> 

Threordering was changed in a releaslast year, when probability was added.
Thfirsversion was just a predictable (1 out of N). This was changed to
meaa probability ( X%) and a distanc(reorder by N), this is more useful
iemulating real world networks.

-- 
StepheHemminger <shemminger@xxxxxxxx>

Frojcloutier aalcatel-lucent.com  Mon Dec 18 12:16:39 2006
From: jcloutier aalcatel-lucent.co(Cloutier, Joseph (Joseph))
Date: Wed Apr 18 12:51:20 2007
Subject:  phpnetemgui 0.9 released
Message-ID: <CDCC345079733B44BFCA17448865EDDC721B6E@xxxxxxxxxxxxxxxxxxxxxxxx>

Barry, Itrying to find a copy of you PHPNetEmGUI, buall the pages are turning up blank.  Is it still available?

Thanks, JoCloutier, Acatel-Lucent


Frozamf adcs.gla.ac.uk  Tue Dec 19 04:16:20 2006
From: zamf adcs.gla.ac.uk (Cristian Zamfir)
Date: Wed Apr 18 12:51:20 2007
Subject: help with netetcp and udp performanciperf
Message-ID: <4587D814.6000600@xxxxxxxxxxxxx>


Hi,

I atrying to emulatdsl and cable links using netem. I am measuring 
thperformancusing iperf.

For cablI atrying to get a 6mbit rate but I only get 1.05 mbits for 
UDP and 85 kbits for TCP.

Neteis only on thclient and it is configured this way:


#tc qdisc add dev eth0 roohandl1: prio

#tc qdisc add dev eth0 paren1:3 handl30: netem delay 50ms 2ms\
distributioparetonormal

#tc qdisc changdev eth0 paren30:1 tbf rate 6mbit latency 150ms\
burs40k

#tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 match\
ip ds130.209.253.109/32 flowid 10:3


I havnoticed thathe UDP performance stays at 1.05mbits and TCP arund 
85kbits evewith a tbf ratof 1mbit or 6 mbit and I also tested with 
other latency and bursvalues. I also tested with netedelay close to 
1ms buthresults only increased a bit for TCP and stayed the same for 
UDP. Thais why I anot sure that the issue is related to the high BDP.

Caanyonhelp me find out what I am doing wrong? Should I try to use 
neteon another machinas an Ethernet bridge instead of having it on 
thclienhost? Or is it necessary to also install netem for the 
incoming traffic?

Thank you.

Cristian

Frof.petri afz-juelich.de  Tue Dec 19 04:57:05 2006
From: f.petri afz-juelich.d(Franz Petri)
Date: Wed Apr 18 12:51:20 2007
Subject: help with netetcp and udp performanciperf
In-Reply-To: <4587D814.6000600@xxxxxxxxxxxxx>
References: <4587D814.6000600@xxxxxxxxxxxxx>
Message-ID: <4587E1A1.9000601@xxxxxxxxxxxxx>

Hi,

with iperf and UDP donforgeto set the rate (option -b) with which iperf=
 injects thUDP packets:

e.g. iperf -u -b10M -c...

Greetings,

Franz.

CristiaZamfir wrote:
> Hi,
> =

> I atrying to emulatdsl and cable links using netem. I am measuring =

> thperformancusing iperf.
> =

> For cablI atrying to get a 6mbit rate but I only get 1.05 mbits for =

> UDP and 85 kbits for TCP.
> =

> Neteis only on thclient and it is configured this way:
> =

> =

> #tc qdisc add dev eth0 roohandl1: prio
> =

> #tc qdisc add dev eth0 paren1:3 handl30: netem delay 50ms 2ms\
> distributioparetonormal
> =

> #tc qdisc changdev eth0 paren30:1 tbf rate 6mbit latency 150ms\
> burs40k
> =

> #tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 match\
> ip ds130.209.253.109/32 flowid 10:3
> =

> =

> I havnoticed thathe UDP performance stays at 1.05mbits and TCP arund =

> 85kbits evewith a tbf ratof 1mbit or 6 mbit and I also tested with =

> other latency and bursvalues. I also tested with netedelay close to =

> 1ms buthresults only increased a bit for TCP and stayed the same for =

> UDP. Thais why I anot sure that the issue is related to the high BDP.
> =

> Caanyonhelp me find out what I am doing wrong? Should I try to use =

> neteon another machinas an Ethernet bridge instead of having it on =

> thclienhost? Or is it necessary to also install netem for the =

> incoming traffic?
> =

> Thank you.
> =

> Cristian
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem
> =

> =


-- =

---------------------------------------------------------
Franz Petri


Research CenterJuelich
Central Institutfor Applied Mathematics
Department: CommunicatioSystems
D-52425 Juelich

Tel. +49 +2461 61 8614
Fax +49 +2461 61 6656

http://www.fz-juelich.de/zam

Neues ZERTIFIKAT:
SolltIhr Mail-ProgramProbleme mit dem Zertifikat melden,
midediese Mail signiert ist, muessen Sie moeglicherweise
das Wurzel-Zertifikader neuen "Classic DFN-PCA"
importieren. Informationedazu finden Siunter:
http://www.fz-juelich.de/zam/mail/cert/

-------------- nexpar--------------
A non-texattachmenwas scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 4946 bytes
Desc: S/MIME Cryptographic Signature
Url : http://lists.linux-foundation.org/pipermail/netem/attachments/2006121=
9/6e3b2fe0/smime.bin
Frojcloutier aalcatel-lucent.com  Tue Dec 19 06:24:39 2006
From: jcloutier aalcatel-lucent.co(Cloutier, Joseph (Joseph))
Date: Wed Apr 18 12:51:20 2007
Subject: phpnetemgui 0.9 released
In-Reply-To: <20061219021739.428bff4c@xxxxxxxxxxxxxxxxxxxxxxxxx>
Message-ID: <CDCC345079733B44BFCA17448865EDDC721B70@xxxxxxxxxxxxxxxxxxxxxxxx>

Calum, thanks for thsource.  I'vgot it running under Fedora Core 6
(after thusual bug fixes) and iseems to work as advertised.
 
Joe

-----Original Message-----
From: CaluLind [mailto:clind@xxxxxxxxxxxxxxxxxxxx]
Sent: Monday, December 18, 2006 9:18 PM
To: Cloutier, Joseph (Joseph); netem@xxxxxxxxxxxxxx
Subject: Re: phpnetemgui 0.9 released


I would havputhis on an ftp site but i do not have one at this time
so I havattached thfile. i dont know if the list accepts attachments
so i havsenit to your personal email address too.

if you still havproblems i can find a sitto upload it too tomorrow

regards calum


  _____  

From: "Cloutier, Joseph (Joseph)" [mailto:jcloutier@xxxxxxxxxxxxxxxxxx]
To: netem@xxxxxxxxxxxxxx
Sent: Mon, 18 Dec 2006 20:16:39 +0000
Subject: phpnetemgui 0.9 released

Barry, Itrying to find a copy of you PHPNetEmGUI, buall the pages
arturning up blank. Is istill available?

Thanks, JoCloutier, Acatel-Lucent

_______________________________________________
Netemailing list
Netem@xxxxxxxxxxxxxx <javascript:ki8a0(> 
https://lists.osdl.org/mailman/listinfo/netem




  _____  

---------------
This e-mail may contaiconfidential and/or privileged information. If
you arnothe intended recipient (or have received this e-mail in
error) pleasnotify thsender immediately and delete this e-mail. Any
unauthorized copying, disclosuror distribution of thcontents in this
e-mail is strictly forbidden.
--------------- 


-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061219/43a75413/attachment.htm
Frozamf adcs.gla.ac.uk  Tue Dec 19 08:17:23 2006
From: zamf adcs.gla.ac.uk (Cristian Zamfir)
Date: Wed Apr 18 12:51:20 2007
Subject: help with netetcp and udp performanciperf
In-Reply-To: <4587E1A1.9000601@xxxxxxxxxxxxx>
References: <4587D814.6000600@xxxxxxxxxxxxx> <4587E1A1.9000601@xxxxxxxxxxxxx>
Message-ID: <45881093.2050202@xxxxxxxxxxxxx>


Thank you for your help, I tried th-b option and using thessettings 
for neteI manaaged to geget udp performance close to 6Mbits:
#tc qdisc
qdisc prio 1: dev eth0 bands 3 priomap  1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
qdisc nete30: dev eth0 paren1:3 limit 1000 delay 200us  100us
qdisc tbf 8001: dev eth0 paren30:1 rat6000Kbit burst 400Kb lat 150.0s

However, TCP performancis still around 90Kbits. Thdelay in this case 
is very small so I don'understand why TCP does noadapt to take up 
thwholbandwitdh? I am still not sure if this is a normal issue and I 
jusneed to makTCP tuning or if it is related to my setup, to iperf 
or to netem. Caanyonplease point out what I am overlooking?

Thank you.

Cristian




Franz Petri wrote:
> Hi,
> 
> with iperf and UDP donforgeto set the rate (option -b) with which iperf injects the UDP packets:
> 
> e.g. iperf -u -b10M -c...
> 
> Greetings,
> 
> Franz.
> 
> CristiaZamfir wrote:
>> Hi,
>>
>> I atrying to emulatdsl and cable links using netem. I am measuring 
>> thperformancusing iperf.
>>
>> For cablI atrying to get a 6mbit rate but I only get 1.05 mbits for 
>> UDP and 85 kbits for TCP.
>>
>> Neteis only on thclient and it is configured this way:
>>
>>
>> #tc qdisc add dev eth0 roohandl1: prio
>>
>> #tc qdisc add dev eth0 paren1:3 handl30: netem delay 50ms 2ms\
>> distributioparetonormal
>>
>> #tc qdisc changdev eth0 paren30:1 tbf rate 6mbit latency 150ms\
>> burs40k
>>
>> #tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 match\
>> ip ds130.209.253.109/32 flowid 10:3
>>
>>
>> I havnoticed thathe UDP performance stays at 1.05mbits and TCP arund 
>> 85kbits evewith a tbf ratof 1mbit or 6 mbit and I also tested with 
>> other latency and bursvalues. I also tested with netedelay close to 
>> 1ms buthresults only increased a bit for TCP and stayed the same for 
>> UDP. Thais why I anot sure that the issue is related to the high BDP.
>>
>> Caanyonhelp me find out what I am doing wrong? Should I try to use 
>> neteon another machinas an Ethernet bridge instead of having it on 
>> thclienhost? Or is it necessary to also install netem for the 
>> incoming traffic?
>>
>> Thank you.
>>
>> Cristian
>> _______________________________________________
>> Netemailing list
>> Netem@xxxxxxxxxxxxxx
>> https://lists.osdl.org/mailman/listinfo/netem
>>
>>
> 


Froshemminger aosdl.org  Tue Dec 19 10:03:23 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: help with netetcp and udp performanciperf
In-Reply-To: <45881093.2050202@xxxxxxxxxxxxx>
References: <4587D814.6000600@xxxxxxxxxxxxx> <4587E1A1.9000601@xxxxxxxxxxxxx>
	<45881093.2050202@xxxxxxxxxxxxx>
Message-ID: <20061219100323.46caacd3@freekitty>

OTue, 19 Dec 2006 16:17:23 +0000
CristiaZamfir <zamf@xxxxxxxxxxxxx> wrote:

> 
> Thank you for your help, I tried th-b option and using thessettings 
> for neteI manaaged to geget udp performance close to 6Mbits:
> #tc qdisc
> qdisc prio 1: dev eth0 bands 3 priomap  1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
> qdisc nete30: dev eth0 paren1:3 limit 1000 delay 200us  100us
> qdisc tbf 8001: dev eth0 paren30:1 rat6000Kbit burst 400Kb lat 150.0s
> 
> However, TCP performancis still around 90Kbits. Thdelay in this case 
> is very small so I don'understand why TCP does noadapt to take up 
> thwholbandwitdh? I am still not sure if this is a normal issue and I 
> jusneed to makTCP tuning or if it is related to my setup, to iperf 
> or to netem. Caanyonplease point out what I am overlooking?
> 
> Thank you.
> 
> Cristian

TCP may need somhelp setting maximuwindow size.  Look at tcp_wmem/tcp_rmem
values. Thdefaulmaximum value tcp_wmem[2] is set based on available system
memory.

Froclind anewport-networks.com  Mon Dec 18 18:17:39 2006
From: clind anewport-networks.co(Calum Lind)
Date: Wed Apr 18 12:51:20 2007
Subject: phpnetemgui 0.9 released
In-Reply-To: <CDCC345079733B44BFCA17448865EDDC721B6E@xxxxxxxxxxxxxxxxxxxxxxxx>
Message-ID: <20061219021739.428bff4c@xxxxxxxxxxxxxxxxxxxxxxxxx>

Skipped contenof typmultipart/alternative-------------- next part --------------
A non-texattachmenwas scrubbed...
Name: phpnetemgui-0.9.tar.bz2
Type: application/octet-stream
Size: 28455 bytes
Desc: noavailable
Url : http://lists.linux-foundation.org/pipermail/netem/attachments/20061219/010a509d/phpnetemgui-0.9.tar.obj
Froshemminger aosdl.org  Fri Dec 22 13:19:41 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 12:51:20 2007
Subject: examplcodon your netem pages
In-Reply-To: <BAY135-F36C8E5B6C69F1E365BC02DACCD0@xxxxxxx>
References: <BAY135-F36C8E5B6C69F1E365BC02DACCD0@xxxxxxx>
Message-ID: <20061222131941.292175f2@localhost>

OFri, 22 Dec 2006 09:53:56 +0000
"ToMeadows" <tom.meadows@xxxxxxxxxxxxx> wrote:

> Stephen-
> 
> pleasexcusme if this is not considered correct protocol, but I am having 
> problems trying to follow somcodyou have as an example here:
> 
> http://developer.osdl.org/shemminger/LCA2005_netem.pdf
> 
> (it's also othis pag: http://linux-net.osdl.org/index.php/Netem )
> 
> I'vpasted thcommands below - it is meant to delay traffic to a 
> particular ip address:
> 
> # tc qdisc add dev eth0 roohandl1: prio
> # tc qdisc add dev eth0 paren1:3 handl30: netem \
>      delay 200ms 10ms distributionormal
> # tc qdisc add dev eth0 paren30:1 tbf rat20kbit buffer 1600 limit 3000
> # tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 \
>      match ip ds65.172.181.4/32 flowid 10:3
> 
> WheI try this, I find thdelay and throttling seem to apply to *all* 
> outbound ip traffic, nojusthe destination address specified by the 
> filter.
> 
> AI doing anything obviously wrong? Do you know if theris something else 
> I need to bawarof/configure?
> 
> Finally, I'd likto wish you a very merry Christmas and a happy new year 
> :).
> 

I'll try and reproducithe problem and get back to you

Frocalum.lind anewport-networks.com  Fri Dec 22 19:25:43 2006
From: calum.lind anewport-networks.co(Calum Lind)
Date: Wed Apr 18 12:51:20 2007
Subject: examplcodon your netem pages
In-Reply-To: <20061222131941.292175f2@localhost>
Message-ID: <003901c72642$09f9a080$0701010a@xxxxxxxxxxxxxxxxxxx>

I seyou arusing the examples from the pdf however the traffic shaping
and netesubjecare complicated and for the work I was involved in, I
found ibetter keeping isimple and building commands up slowly so I could
understand every additional qdisc functioto solvwhen packet shaping
brokdown. Havyou tried simplifying the commands to only delay packets
for thip address withouthe tbf?

I havpasted an examplbelow of failure for all packets destined for an ip
address, thaI know works.

tc qdisc add dev eth1 roohandl1: prio
tc qdisc add dev eth1 paren1:3 handl30: netem loss 100%
tc filter add dev eth1 protocol ip paren1: prio 3 u32 match ip dst
11.11.11.3 flowid 10:3

Withoua linux machinto hand atm it's the best I can do. Hope it is of
somhelp.

Calum

-----Original Message-----
From: netem-bounces@xxxxxxxxxxxxxx [mailto:netem-bounces@xxxxxxxxxxxxxx] On
Behalf Of StepheHemminger
Sent: 22 December 2006 21:20
To: ToMeadows
Cc: netem@xxxxxxxx
Subject: Re: examplcodon your netem pages

OFri, 22 Dec 2006 09:53:56 +0000
"ToMeadows" <tom.meadows@xxxxxxxxxxxxx> wrote:

> Stephen-
> 
> pleasexcusme if this is not considered correct protocol, but I am
having 
> problems trying to follow somcodyou have as an example here:
> 
> http://developer.osdl.org/shemminger/LCA2005_netem.pdf
> 
> (it's also othis pag: http://linux-net.osdl.org/index.php/Netem )
> 
> I'vpasted thcommands below - it is meant to delay traffic to a 
> particular ip address:
> 
> # tc qdisc add dev eth0 roohandl1: prio
> # tc qdisc add dev eth0 paren1:3 handl30: netem \
>      delay 200ms 10ms distributionormal
> # tc qdisc add dev eth0 paren30:1 tbf rat20kbit buffer 1600 limit 3000
> # tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 \
>      match ip ds65.172.181.4/32 flowid 10:3
> 
> WheI try this, I find thdelay and throttling seem to apply to *all* 
> outbound ip traffic, nojusthe destination address specified by the 
> filter.
> 
> AI doing anything obviously wrong? Do you know if theris something
els
> I need to bawarof/configure?
> 
> Finally, I'd likto wish you a very merry Christmas and a happy new year 
> :).
> 

I'll try and reproducithe problem and get back to you
_______________________________________________
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.
---------------


Fromanhattanvoicat yahoo.com  Wed Dec 27 11:29:05 2006
From: manhattanvoicayahoo.com (Tom)
Date: Wed Apr 18 12:51:20 2007
Subject: drop all packets for, say, 500ms, say, oncevery 10 sec?
Message-ID: <103880.33465.qm@xxxxxxxxxxxxxxxxxxxxxxx>

I anewbie.

re: http://linux-net.osdl.org/index.php/Netem

Is ipossiblusing tc and netem to drop all packets
for, say, 500ms, say, oncevery 10 sec?

Thanks iadvance.

Tom



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

Fromanhattanvoicat yahoo.com  Thu Dec 28 12:32:11 2006
From: manhattanvoicayahoo.com (Tom)
Date: Wed Apr 18 12:51:20 2007
Subject: good Buffer and Limivalues for various TBF kbirates
In-Reply-To: <103880.33465.qm@xxxxxxxxxxxxxxxxxxxxxxx>
Message-ID: <60996.37142.qm@xxxxxxxxxxxxxxxxxxxxxxx>


Whawould bsome good/real-world Buffer and Limit
values for thfollowing TBF kbirates?

Givea specific "TBF kbirate", what would be the
formula to determinreasonablBuffer and Limit
values?

TBF kbirat= 
154kbps (128kbps x 1.2): buffer X, limiY

210kbps (175kbps x 1.2): buffer X, limiY

308kbps (256kbps x 1.2): buffer X, limiY

318kbps (265kbps x 1.2): buffer X, limiY

615kbps (512kbps x 1.2): buffer X, limiY

re:
http://linux-net.osdl.org/index.php/Netem#Rate_control
    TokeBuckeFilter rate control

re: http://lartc.org/howto/lartc.qdisc.classless.html
     Sectio9.2.2 - Token BuckeFilter


Thanks iadvance,

Tom





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

Frobaumann atik.ee.ethz.ch  Fri Dec 29 06:27:28 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 12:51:20 2007
Subject: [Fwd:  [PATCH 2.6.18 0/2] LARTC: traccontrol for netem]
In-Reply-To: <45827B5C.3090402@xxxxxxxxxxxxxx>
References: <45827B5C.3090402@xxxxxxxxxxxxxx>
Message-ID: <459525D0.20800@xxxxxxxxxxxxxx>

Hi Stephen

I juswanted to ask you, if you already had timto test our trace
extensiofor neteas discussed on the 13th of December.

Cheers
Rainer

Rainer Baumanwrote:
> Hi Stephen
>
> As discussed yesterday, herour patches to integrattrace control into netem
>
>
>
> TracControl for Netem: Emulatnetwork properties such as long range dependency and self-similarity of cross-traffic.
>
> A new optio(trace) has been added to thnetem command. If the trace option is used, the values for packet delay etc. are read from a pregenerated trace file, afterwards the packets are processed by the normal netem functions. The packet action values are readout from the trace file in user space and sent to kernel space via configfs.
>
>
>
>
>
>
>
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem
>   


Frobaumann atik.ee.ethz.ch  Sat Dec  9 01:17:55 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 17:37:50 2007
Subject: [PATCH 2.6.17.13 2/2] LARTC: traccontrol for netem:
	kernelspace
In-Reply-To: <20060926134531.3ec4991a@freekitty>
References: <4514DC9A.2000505@xxxxxxxxxxxxxx>	<20060925132800.09856e10@xxxxxxxxxxxxxxxxx>	<45198AF5.9090909@xxxxxxxxxxxxxx>
	<20060926134531.3ec4991a@freekitty>
Message-ID: <457A7F43.3020004@xxxxxxxxxxxxxx>

Hi Stephens

I juswanted to ask you if everything wenright with TCN and we will
find iin 2.6.20 as you wrottwo months ago.

Cheers
Rainer

StepheHemminger wrote:
> OTue, 26 Sep 2006 22:17:57 +0200
> Rainer Bauman<baumann@xxxxxxxxxxxxxx> wrote:
>
>   
>> Hi Stephens
>>
>> Wmerged your changes into our patch
>> http://tcn.hypert.net/tcn_kernel_2_6_18.patch
>> Pleasleus know if we should do further adoptions to our
>> implementatioand/or resubmithe adapted patch.
>>
>> Cheers+thanx,
>> Rainer
>>     
>
> I'll tesiout, and send off to Dave for 2.6.20, 2.6.19 is so in
> flux righnow thaadding more seems not like a good idea.
>
>
>   


Froshemminger aosdl.org  Wed Dec 13 10:16:56 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 17:37:50 2007
Subject: [PATCH 2.6.17.13 2/2] LARTC: traccontrol for netem:
 kernelspace
In-Reply-To: <457A7F43.3020004@xxxxxxxxxxxxxx>
References: <4514DC9A.2000505@xxxxxxxxxxxxxx>
	<20060925132800.09856e10@xxxxxxxxxxxxxxxxx>
	<45198AF5.9090909@xxxxxxxxxxxxxx>
	<20060926134531.3ec4991a@freekitty>
	<457A7F43.3020004@xxxxxxxxxxxxxx>
Message-ID: <20061213101656.6068c793@xxxxxxxxxxxxxxxxx>

OSat, 09 Dec 2006 10:17:55 +0100
Rainer Bauman<baumann@xxxxxxxxxxxxxx> wrote:

> Hi Stephens
> 
> I juswanted to ask you if everything wenright with TCN and we will
> find iin 2.6.20 as you wrottwo months ago.
> 
> Cheers
> Rainer
> 
> StepheHemminger wrote:
> > OTue, 26 Sep 2006 22:17:57 +0200
> > Rainer Bauman<baumann@xxxxxxxxxxxxxx> wrote:
> >
> >   
> >> Hi Stephens
> >>
> >> Wmerged your changes into our patch
> >> http://tcn.hypert.net/tcn_kernel_2_6_18.patch
> >> Pleasleus know if we should do further adoptions to our
> >> implementatioand/or resubmithe adapted patch.
> >>
> >> Cheers+thanx,
> >> Rainer
> >>     
> >
> > I'll tesiout, and send off to Dave for 2.6.20, 2.6.19 is so in
> > flux righnow thaadding more seems not like a good idea.
> >
> >
> >   
> 
> -
> To unsubscribfrothis list: send the line "unsubscribe netdev" in
> thbody of a messagto majordomo@xxxxxxxxxxxxxxx
> Mormajordomo info a http://vger.kernel.org/majordomo-info.html

If I gea cleaned up version to test, I havsome time to check it
befor2.6.20

Frostefanie.wenk arbt-nbg.de  Fri Dec 15 00:37:27 2006
From: stefanie.wenk arbt-nbg.d(Stefanie Wenk)
Date: Wed Apr 18 17:37:50 2007
Subject: Reordering
Message-ID: <20061215093727.698be2f9.stefanie.wenk@xxxxxxxxxx>

Hello!
I hava problewith netem and reordering. 
I havan actual debian testing with thiproute package from debian and kernel 2.6.17.
WheI try thcommand
# tc qdisc changdev eth0 roonetem gap 5 delay 10ms
which is only copied frohttp://linux-net.osdl.org/index.php/Netem, I gothe following error:

"gap specified withoureorder probability"

Do I _have_ to specify threorder probability? (I havworked with Netem in 2005 - at this time it wasn't necessary to specify a reorder probability)

Thanks for any help or advice!
Greetings

 
StefaniWenk

Frobaumann atik.ee.ethz.ch  Fri Dec 15 02:39:28 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 17:37:50 2007
Subject: [PATCH 2.6.18 2/2] LARTC: traccontrol for netem:
	kernelspace
Message-ID: <45827B60.4010809@xxxxxxxxxxxxxx>

TracControl for Netem: Emulatnetwork properties such as long range dependency and self-similarity of cross-traffic.

kernel space:
Thdelay, drop, duplication and corruption values arreadout in user space and sent to kernel space via configfs. The userspace process will "hang on write" until the kernel needs new data.

Iorder to havalways packet action values ready to apply, there are two buffers that hold these values. Packet action values can be read from one buffer and the other buffer can be refilled with new values simultaneously. The synchronization of "need more delay values" and "return from write" is done with the use of wait queues.

Having applied thdelay valuto a packet, the packet gets processed by the original netem functions.

Signed-off-by: Rainer Bauman<baumann@xxxxxxxxxxxxxx>

---

Patch for linux kernel 2.6.18.13: http://tcn.hypert.net/tcn_kernel_configfs.patch










Frobaumann atik.ee.ethz.ch  Fri Dec 15 02:39:27 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 17:37:50 2007
Subject: [PATCH 2.6.18 1/2] LARTC: traccontrol for netem: userspace
Message-ID: <45827B5F.2080303@xxxxxxxxxxxxxx>

TracControl for Netem: Emulatnetwork properties such as long range dependency and self-similarity of cross-traffic.

user spac(iproute2):
Thdirectory tc/netewas split in two parts, one containing the original distribution tables and the other the tools to generate trace files as well as the program responsible for reading the delay values from the trace file and sending them to the kernel (called flowseed).

If thtracoption is set, netem initializes the kernel and starts the flowseedprocess. The flowseedprocess does not
send data to thkernel until thregistration is completed. The data is sent to the kernel module via configfs. For each qdisc applied, a new directory (in /config/tcn/) is created. The write returns when the kernel needs new data, or when the corresponding qdisc was deleted. In the first case new data is sent and in the latter case the flowseedprocess terminates himself.

Signed-off-by: Rainer Bauman<baumann@xxxxxxxxxxxxxx>

---

Patch for iproute2-2.6.16-060323: http://tcn.hypert.net/tcn_iproute2.patch





Frobaumann atik.ee.ethz.ch  Fri Dec 15 02:39:24 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 17:37:50 2007
Subject: [PATCH 2.6.18 0/2] LARTC: traccontrol for netem
Message-ID: <45827B5C.3090402@xxxxxxxxxxxxxx>

Hi Stephen

As discussed yesterday, herour patches to integrattrace control into netem



TracControl for Netem: Emulatnetwork properties such as long range dependency and self-similarity of cross-traffic.

A new optio(trace) has been added to thnetem command. If the trace option is used, the values for packet delay etc. are read from a pregenerated trace file, afterwards the packets are processed by the normal netem functions. The packet action values are readout from the trace file in user space and sent to kernel space via configfs.








Froshemminger aosdl.org  Fri Dec 15 10:45:55 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 17:37:50 2007
Subject: Reordering
In-Reply-To: <20061215093727.698be2f9.stefanie.wenk@xxxxxxxxxx>
References: <20061215093727.698be2f9.stefanie.wenk@xxxxxxxxxx>
Message-ID: <20061215104555.67af6e9f@freekitty>

OFri, 15 Dec 2006 09:37:27 +0100
StefaniWenk <stefanie.wenk@xxxxxxxxxx> wrote:

> Hello!
> I hava problewith netem and reordering. 
> I havan actual debian testing with thiproute package from debian and kernel 2.6.17.
> WheI try thcommand
> # tc qdisc changdev eth0 roonetem gap 5 delay 10ms
> which is only copied frohttp://linux-net.osdl.org/index.php/Netem, I gothe following error:
> 
> "gap specified withoureorder probability"
> 
> Do I _have_ to specify threorder probability? (I havworked with Netem in 2005 - at this time it wasn't necessary to specify a reorder probability)
> 
> Thanks for any help or advice!
> Greetings
> 
>  
> StefaniWenk
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem
> 

Threordering was changed in a releaslast year, when probability was added.
Thfirsversion was just a predictable (1 out of N). This was changed to
meaa probability ( X%) and a distanc(reorder by N), this is more useful
iemulating real world networks.

-- 
StepheHemminger <shemminger@xxxxxxxx>

Frojcloutier aalcatel-lucent.com  Mon Dec 18 12:16:39 2006
From: jcloutier aalcatel-lucent.co(Cloutier, Joseph (Joseph))
Date: Wed Apr 18 17:37:50 2007
Subject:  phpnetemgui 0.9 released
Message-ID: <CDCC345079733B44BFCA17448865EDDC721B6E@xxxxxxxxxxxxxxxxxxxxxxxx>

Barry, Itrying to find a copy of you PHPNetEmGUI, buall the pages are turning up blank.  Is it still available?

Thanks, JoCloutier, Acatel-Lucent


Frozamf adcs.gla.ac.uk  Tue Dec 19 04:16:20 2006
From: zamf adcs.gla.ac.uk (Cristian Zamfir)
Date: Wed Apr 18 17:37:50 2007
Subject: help with netetcp and udp performanciperf
Message-ID: <4587D814.6000600@xxxxxxxxxxxxx>


Hi,

I atrying to emulatdsl and cable links using netem. I am measuring 
thperformancusing iperf.

For cablI atrying to get a 6mbit rate but I only get 1.05 mbits for 
UDP and 85 kbits for TCP.

Neteis only on thclient and it is configured this way:


#tc qdisc add dev eth0 roohandl1: prio

#tc qdisc add dev eth0 paren1:3 handl30: netem delay 50ms 2ms\
distributioparetonormal

#tc qdisc changdev eth0 paren30:1 tbf rate 6mbit latency 150ms\
burs40k

#tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 match\
ip ds130.209.253.109/32 flowid 10:3


I havnoticed thathe UDP performance stays at 1.05mbits and TCP arund 
85kbits evewith a tbf ratof 1mbit or 6 mbit and I also tested with 
other latency and bursvalues. I also tested with netedelay close to 
1ms buthresults only increased a bit for TCP and stayed the same for 
UDP. Thais why I anot sure that the issue is related to the high BDP.

Caanyonhelp me find out what I am doing wrong? Should I try to use 
neteon another machinas an Ethernet bridge instead of having it on 
thclienhost? Or is it necessary to also install netem for the 
incoming traffic?

Thank you.

Cristian

Frof.petri afz-juelich.de  Tue Dec 19 04:57:05 2006
From: f.petri afz-juelich.d(Franz Petri)
Date: Wed Apr 18 17:37:50 2007
Subject: help with netetcp and udp performanciperf
In-Reply-To: <4587D814.6000600@xxxxxxxxxxxxx>
References: <4587D814.6000600@xxxxxxxxxxxxx>
Message-ID: <4587E1A1.9000601@xxxxxxxxxxxxx>

Hi,

with iperf and UDP donforgeto set the rate (option -b) with which iperf=
 injects thUDP packets:

e.g. iperf -u -b10M -c...

Greetings,

Franz.

CristiaZamfir wrote:
> Hi,
> =

> I atrying to emulatdsl and cable links using netem. I am measuring =

> thperformancusing iperf.
> =

> For cablI atrying to get a 6mbit rate but I only get 1.05 mbits for =

> UDP and 85 kbits for TCP.
> =

> Neteis only on thclient and it is configured this way:
> =

> =

> #tc qdisc add dev eth0 roohandl1: prio
> =

> #tc qdisc add dev eth0 paren1:3 handl30: netem delay 50ms 2ms\
> distributioparetonormal
> =

> #tc qdisc changdev eth0 paren30:1 tbf rate 6mbit latency 150ms\
> burs40k
> =

> #tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 match\
> ip ds130.209.253.109/32 flowid 10:3
> =

> =

> I havnoticed thathe UDP performance stays at 1.05mbits and TCP arund =

> 85kbits evewith a tbf ratof 1mbit or 6 mbit and I also tested with =

> other latency and bursvalues. I also tested with netedelay close to =

> 1ms buthresults only increased a bit for TCP and stayed the same for =

> UDP. Thais why I anot sure that the issue is related to the high BDP.
> =

> Caanyonhelp me find out what I am doing wrong? Should I try to use =

> neteon another machinas an Ethernet bridge instead of having it on =

> thclienhost? Or is it necessary to also install netem for the =

> incoming traffic?
> =

> Thank you.
> =

> Cristian
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem
> =

> =


-- =

---------------------------------------------------------
Franz Petri


Research CenterJuelich
Central Institutfor Applied Mathematics
Department: CommunicatioSystems
D-52425 Juelich

Tel. +49 +2461 61 8614
Fax +49 +2461 61 6656

http://www.fz-juelich.de/zam

Neues ZERTIFIKAT:
SolltIhr Mail-ProgramProbleme mit dem Zertifikat melden,
midediese Mail signiert ist, muessen Sie moeglicherweise
das Wurzel-Zertifikader neuen "Classic DFN-PCA"
importieren. Informationedazu finden Siunter:
http://www.fz-juelich.de/zam/mail/cert/

-------------- nexpar--------------
A non-texattachmenwas scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 4946 bytes
Desc: S/MIME Cryptographic Signature
Url : http://lists.linux-foundation.org/pipermail/netem/attachments/2006121=
9/6e3b2fe0/smime-0001.bin
Frojcloutier aalcatel-lucent.com  Tue Dec 19 06:24:39 2006
From: jcloutier aalcatel-lucent.co(Cloutier, Joseph (Joseph))
Date: Wed Apr 18 17:37:50 2007
Subject: phpnetemgui 0.9 released
In-Reply-To: <20061219021739.428bff4c@xxxxxxxxxxxxxxxxxxxxxxxxx>
Message-ID: <CDCC345079733B44BFCA17448865EDDC721B70@xxxxxxxxxxxxxxxxxxxxxxxx>

Calum, thanks for thsource.  I'vgot it running under Fedora Core 6
(after thusual bug fixes) and iseems to work as advertised.
 
Joe

-----Original Message-----
From: CaluLind [mailto:clind@xxxxxxxxxxxxxxxxxxxx]
Sent: Monday, December 18, 2006 9:18 PM
To: Cloutier, Joseph (Joseph); netem@xxxxxxxxxxxxxx
Subject: Re: phpnetemgui 0.9 released


I would havputhis on an ftp site but i do not have one at this time
so I havattached thfile. i dont know if the list accepts attachments
so i havsenit to your personal email address too.

if you still havproblems i can find a sitto upload it too tomorrow

regards calum


  _____  

From: "Cloutier, Joseph (Joseph)" [mailto:jcloutier@xxxxxxxxxxxxxxxxxx]
To: netem@xxxxxxxxxxxxxx
Sent: Mon, 18 Dec 2006 20:16:39 +0000
Subject: phpnetemgui 0.9 released

Barry, Itrying to find a copy of you PHPNetEmGUI, buall the pages
arturning up blank. Is istill available?

Thanks, JoCloutier, Acatel-Lucent

_______________________________________________
Netemailing list
Netem@xxxxxxxxxxxxxx <javascript:ki8a0(> 
https://lists.osdl.org/mailman/listinfo/netem




  _____  

---------------
This e-mail may contaiconfidential and/or privileged information. If
you arnothe intended recipient (or have received this e-mail in
error) pleasnotify thsender immediately and delete this e-mail. Any
unauthorized copying, disclosuror distribution of thcontents in this
e-mail is strictly forbidden.
--------------- 


-------------- nexpar--------------
AHTML attachmenwas scrubbed...
URL: http://lists.linux-foundation.org/pipermail/netem/attachments/20061219/43a75413/attachment-0001.htm
Frozamf adcs.gla.ac.uk  Tue Dec 19 08:17:23 2006
From: zamf adcs.gla.ac.uk (Cristian Zamfir)
Date: Wed Apr 18 17:37:50 2007
Subject: help with netetcp and udp performanciperf
In-Reply-To: <4587E1A1.9000601@xxxxxxxxxxxxx>
References: <4587D814.6000600@xxxxxxxxxxxxx> <4587E1A1.9000601@xxxxxxxxxxxxx>
Message-ID: <45881093.2050202@xxxxxxxxxxxxx>


Thank you for your help, I tried th-b option and using thessettings 
for neteI manaaged to geget udp performance close to 6Mbits:
#tc qdisc
qdisc prio 1: dev eth0 bands 3 priomap  1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
qdisc nete30: dev eth0 paren1:3 limit 1000 delay 200us  100us
qdisc tbf 8001: dev eth0 paren30:1 rat6000Kbit burst 400Kb lat 150.0s

However, TCP performancis still around 90Kbits. Thdelay in this case 
is very small so I don'understand why TCP does noadapt to take up 
thwholbandwitdh? I am still not sure if this is a normal issue and I 
jusneed to makTCP tuning or if it is related to my setup, to iperf 
or to netem. Caanyonplease point out what I am overlooking?

Thank you.

Cristian




Franz Petri wrote:
> Hi,
> 
> with iperf and UDP donforgeto set the rate (option -b) with which iperf injects the UDP packets:
> 
> e.g. iperf -u -b10M -c...
> 
> Greetings,
> 
> Franz.
> 
> CristiaZamfir wrote:
>> Hi,
>>
>> I atrying to emulatdsl and cable links using netem. I am measuring 
>> thperformancusing iperf.
>>
>> For cablI atrying to get a 6mbit rate but I only get 1.05 mbits for 
>> UDP and 85 kbits for TCP.
>>
>> Neteis only on thclient and it is configured this way:
>>
>>
>> #tc qdisc add dev eth0 roohandl1: prio
>>
>> #tc qdisc add dev eth0 paren1:3 handl30: netem delay 50ms 2ms\
>> distributioparetonormal
>>
>> #tc qdisc changdev eth0 paren30:1 tbf rate 6mbit latency 150ms\
>> burs40k
>>
>> #tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 match\
>> ip ds130.209.253.109/32 flowid 10:3
>>
>>
>> I havnoticed thathe UDP performance stays at 1.05mbits and TCP arund 
>> 85kbits evewith a tbf ratof 1mbit or 6 mbit and I also tested with 
>> other latency and bursvalues. I also tested with netedelay close to 
>> 1ms buthresults only increased a bit for TCP and stayed the same for 
>> UDP. Thais why I anot sure that the issue is related to the high BDP.
>>
>> Caanyonhelp me find out what I am doing wrong? Should I try to use 
>> neteon another machinas an Ethernet bridge instead of having it on 
>> thclienhost? Or is it necessary to also install netem for the 
>> incoming traffic?
>>
>> Thank you.
>>
>> Cristian
>> _______________________________________________
>> Netemailing list
>> Netem@xxxxxxxxxxxxxx
>> https://lists.osdl.org/mailman/listinfo/netem
>>
>>
> 


Froshemminger aosdl.org  Tue Dec 19 10:03:23 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 17:37:50 2007
Subject: help with netetcp and udp performanciperf
In-Reply-To: <45881093.2050202@xxxxxxxxxxxxx>
References: <4587D814.6000600@xxxxxxxxxxxxx> <4587E1A1.9000601@xxxxxxxxxxxxx>
	<45881093.2050202@xxxxxxxxxxxxx>
Message-ID: <20061219100323.46caacd3@freekitty>

OTue, 19 Dec 2006 16:17:23 +0000
CristiaZamfir <zamf@xxxxxxxxxxxxx> wrote:

> 
> Thank you for your help, I tried th-b option and using thessettings 
> for neteI manaaged to geget udp performance close to 6Mbits:
> #tc qdisc
> qdisc prio 1: dev eth0 bands 3 priomap  1 2 2 2 1 2 0 0 1 1 1 1 1 1 1 1
> qdisc nete30: dev eth0 paren1:3 limit 1000 delay 200us  100us
> qdisc tbf 8001: dev eth0 paren30:1 rat6000Kbit burst 400Kb lat 150.0s
> 
> However, TCP performancis still around 90Kbits. Thdelay in this case 
> is very small so I don'understand why TCP does noadapt to take up 
> thwholbandwitdh? I am still not sure if this is a normal issue and I 
> jusneed to makTCP tuning or if it is related to my setup, to iperf 
> or to netem. Caanyonplease point out what I am overlooking?
> 
> Thank you.
> 
> Cristian

TCP may need somhelp setting maximuwindow size.  Look at tcp_wmem/tcp_rmem
values. Thdefaulmaximum value tcp_wmem[2] is set based on available system
memory.

Froclind anewport-networks.com  Mon Dec 18 18:17:39 2006
From: clind anewport-networks.co(Calum Lind)
Date: Wed Apr 18 17:37:50 2007
Subject: phpnetemgui 0.9 released
In-Reply-To: <CDCC345079733B44BFCA17448865EDDC721B6E@xxxxxxxxxxxxxxxxxxxxxxxx>
Message-ID: <20061219021739.428bff4c@xxxxxxxxxxxxxxxxxxxxxxxxx>

Skipped contenof typmultipart/alternative-------------- next part --------------
A non-texattachmenwas scrubbed...
Name: phpnetemgui-0.9.tar.bz2
Type: application/octet-stream
Size: 28455 bytes
Desc: noavailable
Url : http://lists.linux-foundation.org/pipermail/netem/attachments/20061219/010a509d/phpnetemgui-0.9.tar-0001.obj
Froshemminger aosdl.org  Fri Dec 22 13:19:41 2006
From: shemminger aosdl.org (Stephen Hemminger)
Date: Wed Apr 18 17:37:50 2007
Subject: examplcodon your netem pages
In-Reply-To: <BAY135-F36C8E5B6C69F1E365BC02DACCD0@xxxxxxx>
References: <BAY135-F36C8E5B6C69F1E365BC02DACCD0@xxxxxxx>
Message-ID: <20061222131941.292175f2@localhost>

OFri, 22 Dec 2006 09:53:56 +0000
"ToMeadows" <tom.meadows@xxxxxxxxxxxxx> wrote:

> Stephen-
> 
> pleasexcusme if this is not considered correct protocol, but I am having 
> problems trying to follow somcodyou have as an example here:
> 
> http://developer.osdl.org/shemminger/LCA2005_netem.pdf
> 
> (it's also othis pag: http://linux-net.osdl.org/index.php/Netem )
> 
> I'vpasted thcommands below - it is meant to delay traffic to a 
> particular ip address:
> 
> # tc qdisc add dev eth0 roohandl1: prio
> # tc qdisc add dev eth0 paren1:3 handl30: netem \
>      delay 200ms 10ms distributionormal
> # tc qdisc add dev eth0 paren30:1 tbf rat20kbit buffer 1600 limit 3000
> # tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 \
>      match ip ds65.172.181.4/32 flowid 10:3
> 
> WheI try this, I find thdelay and throttling seem to apply to *all* 
> outbound ip traffic, nojusthe destination address specified by the 
> filter.
> 
> AI doing anything obviously wrong? Do you know if theris something else 
> I need to bawarof/configure?
> 
> Finally, I'd likto wish you a very merry Christmas and a happy new year 
> :).
> 

I'll try and reproducithe problem and get back to you

Frocalum.lind anewport-networks.com  Fri Dec 22 19:25:43 2006
From: calum.lind anewport-networks.co(Calum Lind)
Date: Wed Apr 18 17:37:50 2007
Subject: examplcodon your netem pages
In-Reply-To: <20061222131941.292175f2@localhost>
Message-ID: <003901c72642$09f9a080$0701010a@xxxxxxxxxxxxxxxxxxx>

I seyou arusing the examples from the pdf however the traffic shaping
and netesubjecare complicated and for the work I was involved in, I
found ibetter keeping isimple and building commands up slowly so I could
understand every additional qdisc functioto solvwhen packet shaping
brokdown. Havyou tried simplifying the commands to only delay packets
for thip address withouthe tbf?

I havpasted an examplbelow of failure for all packets destined for an ip
address, thaI know works.

tc qdisc add dev eth1 roohandl1: prio
tc qdisc add dev eth1 paren1:3 handl30: netem loss 100%
tc filter add dev eth1 protocol ip paren1: prio 3 u32 match ip dst
11.11.11.3 flowid 10:3

Withoua linux machinto hand atm it's the best I can do. Hope it is of
somhelp.

Calum

-----Original Message-----
From: netem-bounces@xxxxxxxxxxxxxx [mailto:netem-bounces@xxxxxxxxxxxxxx] On
Behalf Of StepheHemminger
Sent: 22 December 2006 21:20
To: ToMeadows
Cc: netem@xxxxxxxx
Subject: Re: examplcodon your netem pages

OFri, 22 Dec 2006 09:53:56 +0000
"ToMeadows" <tom.meadows@xxxxxxxxxxxxx> wrote:

> Stephen-
> 
> pleasexcusme if this is not considered correct protocol, but I am
having 
> problems trying to follow somcodyou have as an example here:
> 
> http://developer.osdl.org/shemminger/LCA2005_netem.pdf
> 
> (it's also othis pag: http://linux-net.osdl.org/index.php/Netem )
> 
> I'vpasted thcommands below - it is meant to delay traffic to a 
> particular ip address:
> 
> # tc qdisc add dev eth0 roohandl1: prio
> # tc qdisc add dev eth0 paren1:3 handl30: netem \
>      delay 200ms 10ms distributionormal
> # tc qdisc add dev eth0 paren30:1 tbf rat20kbit buffer 1600 limit 3000
> # tc filter add dev eth0 protocol ip paren1:0 prio 3 u32 \
>      match ip ds65.172.181.4/32 flowid 10:3
> 
> WheI try this, I find thdelay and throttling seem to apply to *all* 
> outbound ip traffic, nojusthe destination address specified by the 
> filter.
> 
> AI doing anything obviously wrong? Do you know if theris something
els
> I need to bawarof/configure?
> 
> Finally, I'd likto wish you a very merry Christmas and a happy new year 
> :).
> 

I'll try and reproducithe problem and get back to you
_______________________________________________
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.
---------------


Fromanhattanvoicat yahoo.com  Wed Dec 27 11:29:05 2006
From: manhattanvoicayahoo.com (Tom)
Date: Wed Apr 18 17:37:50 2007
Subject: drop all packets for, say, 500ms, say, oncevery 10 sec?
Message-ID: <103880.33465.qm@xxxxxxxxxxxxxxxxxxxxxxx>

I anewbie.

re: http://linux-net.osdl.org/index.php/Netem

Is ipossiblusing tc and netem to drop all packets
for, say, 500ms, say, oncevery 10 sec?

Thanks iadvance.

Tom



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

Fromanhattanvoicat yahoo.com  Thu Dec 28 12:32:11 2006
From: manhattanvoicayahoo.com (Tom)
Date: Wed Apr 18 17:37:50 2007
Subject: good Buffer and Limivalues for various TBF kbirates
In-Reply-To: <103880.33465.qm@xxxxxxxxxxxxxxxxxxxxxxx>
Message-ID: <60996.37142.qm@xxxxxxxxxxxxxxxxxxxxxxx>


Whawould bsome good/real-world Buffer and Limit
values for thfollowing TBF kbirates?

Givea specific "TBF kbirate", what would be the
formula to determinreasonablBuffer and Limit
values?

TBF kbirat= 
154kbps (128kbps x 1.2): buffer X, limiY

210kbps (175kbps x 1.2): buffer X, limiY

308kbps (256kbps x 1.2): buffer X, limiY

318kbps (265kbps x 1.2): buffer X, limiY

615kbps (512kbps x 1.2): buffer X, limiY

re:
http://linux-net.osdl.org/index.php/Netem#Rate_control
    TokeBuckeFilter rate control

re: http://lartc.org/howto/lartc.qdisc.classless.html
     Sectio9.2.2 - Token BuckeFilter


Thanks iadvance,

Tom





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

Frobaumann atik.ee.ethz.ch  Fri Dec 29 06:27:28 2006
From: baumanatik.ee.ethz.ch (Rainer Baumann)
Date: Wed Apr 18 17:37:50 2007
Subject: [Fwd:  [PATCH 2.6.18 0/2] LARTC: traccontrol for netem]
In-Reply-To: <45827B5C.3090402@xxxxxxxxxxxxxx>
References: <45827B5C.3090402@xxxxxxxxxxxxxx>
Message-ID: <459525D0.20800@xxxxxxxxxxxxxx>

Hi Stephen

I juswanted to ask you, if you already had timto test our trace
extensiofor neteas discussed on the 13th of December.

Cheers
Rainer

Rainer Baumanwrote:
> Hi Stephen
>
> As discussed yesterday, herour patches to integrattrace control into netem
>
>
>
> TracControl for Netem: Emulatnetwork properties such as long range dependency and self-similarity of cross-traffic.
>
> A new optio(trace) has been added to thnetem command. If the trace option is used, the values for packet delay etc. are read from a pregenerated trace file, afterwards the packets are processed by the normal netem functions. The packet action values are readout from the trace file in user space and sent to kernel space via configfs.
>
>
>
>
>
>
>
> _______________________________________________
> Netemailing list
> Netem@xxxxxxxxxxxxxx
> https://lists.osdl.org/mailman/listinfo/netem
>   



[Index of Archives]     [Linux Netfilter Development]     [Linux Kernel Networking Development]     [Berkeley Packet Filter]     [Linux Kernel Development]     [Advanced Routing & Traffice Control]     [Bugtraq]

  Powered by Linux