[Codel] RRUL Test

[Codel] RRUL Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 353 bytes --]

Hi

Anyone please tell me what are requirements for taking this test:

1) About Operating System :

which os is best-  windows or ubuntu?
can i use windows?

2) can i test it on home network for bufferbloat?
or should i look for some wider network?

3) Any other link explaining about its procedure from scratch because i am
newbie to all this.

Thanks

[-- Attachment #2: Type: text/html, Size: 1112 bytes --]

Re: [Codel] RRUL Test

[Andrew McGregor]

It only runs on some Linux variant.  It is interesting to test on any network.

On Fri, Apr 17, 2015 at 3:22 PM, sahil grover <sahilgrover013@gmail.com> wrote:
> Hi
>
> Anyone please tell me what are requirements for taking this test:
>
> 1) About Operating System :
>
> which os is best-  windows or ubuntu?
> can i use windows?
>
> 2) can i test it on home network for bufferbloat?
> or should i look for some wider network?
>
> 3) Any other link explaining about its procedure from scratch because i am
> newbie to all this.
>
> Thanks
>
> _______________________________________________
> Codel mailing list
> Codel@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/codel
>

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 1105 bytes --]

On Fri, Apr 17, 2015 at 12:18 PM, Andrew McGregor <andrewmcgr@gmail.com>
 wrote:

> It only runs on some Linux variant.  It is interesting to test on any
> network.




will ubuntu 12.04 be ok?

can you please tell me step by step installation of RRUL(or
netperf-wrapper)?



On Fri, Apr 17, 2015 at 12:18 PM, Andrew McGregor <andrewmcgr@gmail.com>
wrote:

> It only runs on some Linux variant.  It is interesting to test on any
> network.
>
> On Fri, Apr 17, 2015 at 3:22 PM, sahil grover <sahilgrover013@gmail.com>
> wrote:
> > Hi
> >
> > Anyone please tell me what are requirements for taking this test:
> >
> > 1) About Operating System :
> >
> > which os is best-  windows or ubuntu?
> > can i use windows?
> >
> > 2) can i test it on home network for bufferbloat?
> > or should i look for some wider network?
> >
> > 3) Any other link explaining about its procedure from scratch because i
> am
> > newbie to all this.
> >
> > Thanks
> >
> > _______________________________________________
> > Codel mailing list
> > Codel@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/codel
> >
>

[-- Attachment #2: Type: text/html, Size: 2109 bytes --]

Re: [Codel] RRUL Test

[Andrew McGregor]

Check the dependencies; if you have the right version of netperf
(later than 2.6.0), it should be OK, but you really ought to be
running a more current distribution with the latest version of
everything relevant for your results to be particularly meaningful.
Ubuntu 14.04 would be your best bet for a distribution that should
just work.

As for step-by-step instructions, sorry, that would take quite some
time to write.

On Fri, Apr 17, 2015 at 7:04 PM, sahil grover <sahilgrover013@gmail.com> wrote:
> On Fri, Apr 17, 2015 at 12:18 PM, Andrew McGregor <andrewmcgr@gmail.com>
> wrote:
>>
>> It only runs on some Linux variant.  It is interesting to test on any
>> network.
>
>
>
>
> will ubuntu 12.04 be ok?
>
> can you please tell me step by step installation of RRUL(or
> netperf-wrapper)?
>
>
>
> On Fri, Apr 17, 2015 at 12:18 PM, Andrew McGregor <andrewmcgr@gmail.com>
> wrote:
>>
>> It only runs on some Linux variant.  It is interesting to test on any
>> network.
>>
>> On Fri, Apr 17, 2015 at 3:22 PM, sahil grover <sahilgrover013@gmail.com>
>> wrote:
>> > Hi
>> >
>> > Anyone please tell me what are requirements for taking this test:
>> >
>> > 1) About Operating System :
>> >
>> > which os is best-  windows or ubuntu?
>> > can i use windows?
>> >
>> > 2) can i test it on home network for bufferbloat?
>> > or should i look for some wider network?
>> >
>> > 3) Any other link explaining about its procedure from scratch because i
>> > am
>> > newbie to all this.
>> >
>> > Thanks
>> >
>> > _______________________________________________
>> > Codel mailing list
>> > Codel@lists.bufferbloat.net
>> > https://lists.bufferbloat.net/listinfo/codel
>> >
>
>

Re: [Codel] RRUL Test

[Toke Høiland-Jørgensen]

sahil grover <sahilgrover013@gmail.com> writes:

> will ubuntu 12.04 be ok?
>
> can you please tell me step by step installation of RRUL(or
> netperf-wrapper)?

There's an ubuntu 12.04 repository with netperf-wrapper and the required
version of netperf at the open build service repository. See:

https://software.opensuse.org/download.html?project=home:tohojo:netperf-wrapper&package=netperf-wrapper

That link contains instructions for how to enable the repository on
Ubuntu and install the netperf-wrapper package.

As far as running it, have a look at the Quick Start in the README
(https://github.com/tohojo/netperf-wrapper#quick-start) or read the man
page.

-Toke

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 1499 bytes --]

Thanks Andrew Sir and Toke Sir for replying.

Toke Sir,

Please clear one more thing.

According to  Quick Start(
https://github.com/tohojo/netperf-wrapper#quick-start):
#### You must install netperf on two computers.Netperf needs to be started
in “server mode” on one #computer so it can listen for commands from the
other.

####Note: You may also use the netserver that is running on
netperf.bufferbloat.net - use “-H #netperf.bufferbloat.net” in the commands
above.


if i choose  "netperf.bufferbloat.net" as a server then i think i can
proceed with one computer for bufferbloat test.
Does it make any difference?
i just need your confirmation.



On Fri, Apr 17, 2015 at 6:12 PM, Toke Høiland-Jørgensen <toke@toke.dk>
wrote:

> sahil grover <sahilgrover013@gmail.com> writes:
>
> > will ubuntu 12.04 be ok?
> >
> > can you please tell me step by step installation of RRUL(or
> > netperf-wrapper)?
>
> There's an ubuntu 12.04 repository with netperf-wrapper and the required
> version of netperf at the open build service repository. See:
>
>
> https://software.opensuse.org/download.html?project=home:tohojo:netperf-wrapper&package=netperf-wrapper
>
> That link contains instructions for how to enable the repository on
> Ubuntu and install the netperf-wrapper package.
>
> As far as running it, have a look at the Quick Start in the README
> (https://github.com/tohojo/netperf-wrapper#quick-start) or read the man
> page.
>
> -Toke
>

[-- Attachment #2: Type: text/html, Size: 4248 bytes --]

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 1891 bytes --]

Hi again

i was searching around and i found i need to install netperf before
netperf-wrapper.
Am i right?
Sorry for asking these type of questions.
But i have to ....
i really need your help.
if possible please reply.




On Fri, Apr 17, 2015 at 8:22 PM, sahil grover <sahilgrover013@gmail.com>
wrote:

> Thanks Andrew Sir and Toke Sir for replying.
>
> Toke Sir,
>
> Please clear one more thing.
>
> According to  Quick Start(
> https://github.com/tohojo/netperf-wrapper#quick-start):
> #### You must install netperf on two computers.Netperf needs to be
> started in “server mode” on one #computer so it can listen for commands
> from the other.
>
> ####Note: You may also use the netserver that is running on
> netperf.bufferbloat.net - use “-H #netperf.bufferbloat.net” in the
> commands above.
>
>
> if i choose  "netperf.bufferbloat.net" as a server then i think i can
> proceed with one computer for bufferbloat test.
> Does it make any difference?
> i just need your confirmation.
>
>
>
> On Fri, Apr 17, 2015 at 6:12 PM, Toke Høiland-Jørgensen <toke@toke.dk>
> wrote:
>
>> sahil grover <sahilgrover013@gmail.com> writes:
>>
>> > will ubuntu 12.04 be ok?
>> >
>> > can you please tell me step by step installation of RRUL(or
>> > netperf-wrapper)?
>>
>> There's an ubuntu 12.04 repository with netperf-wrapper and the required
>> version of netperf at the open build service repository. See:
>>
>>
>> https://software.opensuse.org/download.html?project=home:tohojo:netperf-wrapper&package=netperf-wrapper
>>
>> That link contains instructions for how to enable the repository on
>> Ubuntu and install the netperf-wrapper package.
>>
>> As far as running it, have a look at the Quick Start in the README
>> (https://github.com/tohojo/netperf-wrapper#quick-start) or read the man
>> page.
>>
>> -Toke
>>
>
>

[-- Attachment #2: Type: text/html, Size: 5055 bytes --]

Re: [Codel] RRUL Test

[Rick Jones]

On 04/17/2015 08:49 AM, sahil grover wrote:
> Hi again
>
> i was searching around and i found i need to install netperf before
> netperf-wrapper.
> Am i right?


Yes, you have to install netperf before you can use netperf wrapper.

rick jones

Re: [Codel] RRUL Test

[Jonathan Morton]

> On 17 Apr, 2015, at 17:52, sahil grover <sahilgrover013@gmail.com> wrote:
> 
> if i choose  "netperf.bufferbloat.net" as a server then i think i can proceed with one computer for bufferbloat test.
> Does it make any difference?

Just be aware of which equipment you’re effectively testing for bufferbloat.  A server based elsewhere on the Internet will mean that the test traffic goes through your ISP and your modem, which will probably hold the bottleneck queues for downstream and upstream respectively.  By using two of your own computers, you would instead test an individual piece of equipment such as a router placed between them, or the NICs of the two machines themselves.

 - Jonathan Morton

Re: [Codel] RRUL Test

[Toke Høiland-Jørgensen]

Rick Jones <rick.jones2@hp.com> writes:

>> i was searching around and i found i need to install netperf before
>> netperf-wrapper. Am i right?
>
> Yes, you have to install netperf before you can use netperf wrapper.

Note, however, that if you use the buildservice repositories I linked
previously, the netperf-wrapper package will depend on netperf, so it
should be automatically installed. The repo also contains an updated
version of netperf compiled with the right options. So you *should* be
good to go after that installation step. :)

-Toke

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1.1: Type: text/plain, Size: 2717 bytes --]

Thanks everyone for replying.

i  have followed the procedure told by Toke Sir as in 17-April post.

For running, when i did:

 1. netperf-wrapper rrul netperf.bufferbloat.net

It shows

Warning: Command produced no valid data.
Data series: TCP upload BE
Runner: NetperfDemoRunner
Command: /usr/bin/netperf -P 0 -v 0 -D -0.20 -4 -Y CS0,CS0 -H
netperf.bufferbloat.net -t TCP_STREAM -l 60 -f m   --   -H
netperf.bufferbloat.net
Standard error output:

Warning: Command produced no valid data.
Data series: TCP upload BK
Runner: NetperfDemoRunner
Command: /usr/bin/netperf -P 0 -v 0 -D -0.20 -4 -Y CS1,CS1 -H
netperf.bufferbloat.net -t TCP_STREAM -l 60 -f m   --   -H
netperf.bufferbloat.net
Standard error output:

Warning: Command produced no valid data.
Data series: TCP upload CS5
Runner: NetperfDemoRunner
Command: /usr/bin/netperf -P 0 -v 0 -D -0.20 -4 -Y CS5,CS5 -H
netperf.bufferbloat.net -t TCP_STREAM -l 60 -f m   --   -H
netperf.bufferbloat.net
Standard error output:

Warning: Command produced no valid data.
Data series: TCP upload EF
Runner: NetperfDemoRunner
Command: /usr/bin/netperf -P 0 -v 0 -D -0.20 -4 -Y EF,EF -H
netperf.bufferbloat.net -t TCP_STREAM -l 60 -f m   --   -H
netperf.bufferbloat.net
Standard error output:

Warning: Command produced no valid data.
Data series: Ping (ms) UDP BE
Runner: NetperfDemoRunner
Command: /usr/bin/netperf -P 0 -v 0 -D -0.20 -4  -H netperf.bufferbloat.net
-t UDP_RR -l 70    --   -H netperf.bufferbloat.net
Standard error output:

No output formatter selected.
Test data is in ./rrul-2015-04-20T165921.383607.json.gz (use with -i to
format).



And the output(result.png) is attached.
in output why 2nd chart didn't show anything.






Also tried with (https://github.com/tohojo/netperf-wrapper#quick-start)

2.  netperf-wrapper rrul -p all_scaled -l 60 -H netperf.bufferbloat.net

But in this case it does not even complete. Keeps running
but graph is made



Anybody please help me in correcting this error


On Sat, Apr 18, 2015 at 6:12 PM, Toke Høiland-Jørgensen <toke@toke.dk>
wrote:

> Rick Jones <rick.jones2@hp.com> writes:
>
> >> i was searching around and i found i need to install netperf before
> >> netperf-wrapper. Am i right?
> >
> > Yes, you have to install netperf before you can use netperf wrapper.
>
> Note, however, that if you use the buildservice repositories I linked
> previously, the netperf-wrapper package will depend on netperf, so it
> should be automatically installed. The repo also contains an updated
> version of netperf compiled with the right options. So you *should* be
> good to go after that installation step. :)
>
> -Toke
>

[-- Attachment #1.2: Type: text/html, Size: 4483 bytes --]

[-- Attachment #2: result.png --]
[-- Type: image/png, Size: 147425 bytes --]

Re: [Codel] RRUL Test

[Toke Høiland-Jørgensen]

sahil grover <sahilgrover013@gmail.com> writes:

> i have followed the procedure told by Toke Sir as in 17-April post.
>
> For running, when i did:
>
> 1. netperf-wrapper rrul netperf.bufferbloat.net

Well, it looks like for some reason the upload flows are not working.
What is the output of running:

netperf -D -0.20 -H netperf.bufferbloat.net -t TCP_STREAM

?

-Toke

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 876 bytes --]

By          netperf -D -0.20 -H netperf.bufferbloat.net -t TCP_STREAM


MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to
atl.richb-hanover.com () port 0 AF_INET : demo
Recv   Send    Send
Socket Socket  Message  Elapsed
Size   Size    Size     Time     Throughput
bytes  bytes   bytes    secs.    10^6bits/sec

174760  16384  16384    14.02       0.00

On Mon, Apr 20, 2015 at 8:56 PM, Toke Høiland-Jørgensen <toke@toke.dk>
wrote:

> sahil grover <sahilgrover013@gmail.com> writes:
>
> > i have followed the procedure told by Toke Sir as in 17-April post.
> >
> > For running, when i did:
> >
> > 1. netperf-wrapper rrul netperf.bufferbloat.net
>
> Well, it looks like for some reason the upload flows are not working.
> What is the output of running:
>
> netperf -D -0.20 -H netperf.bufferbloat.net -t TCP_STREAM
>
> ?
>
> -Toke
>

[-- Attachment #2: Type: text/html, Size: 1910 bytes --]

Re: [Codel] RRUL Test

[Toke Høiland-Jørgensen]

sahil grover <sahilgrover013@gmail.com> writes:

> By netperf -D -0.20 -H netperf.bufferbloat.net -t TCP_STREAM
>
> MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to atl.richb-hanover.com
> () port 0 AF_INET : demo
> Recv Send Send 
> Socket Socket Message Elapsed 
> Size Size Size Time Throughput 
> bytes bytes bytes secs. 10^6bits/sec 
>
> 174760 16384 16384 14.02 0.00 

Hmm, it seems something is blocking your upstream TCP flows. Not sure
what can cause this, since the downstream ones seem to work. What
happens if you run this against a different server? E.g.

netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM

-Toke

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 1160 bytes --]

netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM

Shows:


MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to kau.toke.dk ()
port 0 AF_INET : demo
Recv   Send    Send
Socket Socket  Message  Elapsed
Size   Size    Size     Time     Throughput
bytes  bytes   bytes    secs.    10^6bits/sec

 87380  16384  16384    14.00       0.00




On Tue, Apr 21, 2015 at 2:27 PM, Toke Høiland-Jørgensen <toke@toke.dk>
wrote:

> sahil grover <sahilgrover013@gmail.com> writes:
>
> > By netperf -D -0.20 -H netperf.bufferbloat.net -t TCP_STREAM
> >
> > MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to
> atl.richb-hanover.com
> > () port 0 AF_INET : demo
> > Recv Send Send
> > Socket Socket Message Elapsed
> > Size Size Size Time Throughput
> > bytes bytes bytes secs. 10^6bits/sec
> >
> > 174760 16384 16384 14.02 0.00
>
> Hmm, it seems something is blocking your upstream TCP flows. Not sure
> what can cause this, since the downstream ones seem to work. What
> happens if you run this against a different server? E.g.
>
> netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM
>
> -Toke
>

[-- Attachment #2: Type: text/html, Size: 2518 bytes --]

Re: [Codel] RRUL Test

[Toke Høiland-Jørgensen]

sahil grover <sahilgrover013@gmail.com> writes:

> netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM
>
> Shows:
>
> MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to kau.toke.dk () port 0
> AF_INET : demo
> Recv Send Send 
> Socket Socket Message Elapsed 
> Size Size Size Time Throughput 
> bytes bytes bytes secs. 10^6bits/sec 
>
> 87380 16384 16384 14.00 0.00

Right, so there's definitely something messing with the outgoing
traffic. I have no idea what could cause netperf to only be able to send
traffic in one direction (perhaps Rick knows?). Do you have any sort of
firewall between you and the internet that might interfere with traffic?

-Toke

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 3072 bytes --]

No,Firewall is disabled.

tried with wired network

1. netperf -D -0.20 -H netperf.bufferbloat.net -t TCP_STREAM

MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to
atl.richb-hanover.com () port 0 AF_INET : demo
Interim result:    0.40 10^6bits/s over 0.324 seconds ending at
1429627638.283
Interim result:    0.28 10^6bits/s over 0.462 seconds ending at
1429627638.745
Interim result:    1.61 10^6bits/s over 0.244 seconds ending at
1429627638.989
Interim result:    0.98 10^6bits/s over 0.401 seconds ending at
1429627639.390
Interim result:    0.51 10^6bits/s over 1.034 seconds ending at
1429627640.424
Interim result:    0.44 10^6bits/s over 1.491 seconds ending at
1429627641.914
Interim result:    0.27 10^6bits/s over 2.418 seconds ending at
1429627644.333
Interim result:    0.70 10^6bits/s over 1.308 seconds ending at
1429627645.641
Interim result:    0.49 10^6bits/s over 1.334 seconds ending at
1429627646.975
Interim result:    0.40 10^6bits/s over 0.984 seconds ending at
1429627647.959
Recv   Send    Send
Socket Socket  Message  Elapsed
Size   Size    Size     Time     Throughput
bytes  bytes   bytes    secs.    10^6bits/sec

174760  16384  16384    13.07       0.38




2.  netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM


MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to kau.toke.dk ()
port 0 AF_INET : demo
Interim result:    0.53 10^6bits/s over 0.247 seconds ending at
1429627754.173
Interim result:    0.30 10^6bits/s over 0.437 seconds ending at
1429627754.610
Interim result:    1.70 10^6bits/s over 0.462 seconds ending at
1429627755.072
Interim result:    1.24 10^6bits/s over 0.424 seconds ending at
1429627755.496
Interim result:    0.48 10^6bits/s over 1.378 seconds ending at
1429627756.874
Interim result:    0.39 10^6bits/s over 2.686 seconds ending at
1429627759.560
Interim result:    0.47 10^6bits/s over 2.215 seconds ending at
1429627761.775
Interim result:    0.43 10^6bits/s over 2.151 seconds ending at
1429627763.926
Recv   Send    Send
Socket Socket  Message  Elapsed
Size   Size    Size     Time     Throughput
bytes  bytes   bytes    secs.    10^6bits/sec

 87380  16384  16384    12.81       0.42





On Tue, Apr 21, 2015 at 7:00 PM, Toke Høiland-Jørgensen <toke@toke.dk>
wrote:

> sahil grover <sahilgrover013@gmail.com> writes:
>
> > netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM
> >
> > Shows:
> >
> > MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to kau.toke.dk
> () port 0
> > AF_INET : demo
> > Recv Send Send
> > Socket Socket Message Elapsed
> > Size Size Size Time Throughput
> > bytes bytes bytes secs. 10^6bits/sec
> >
> > 87380 16384 16384 14.00 0.00
>
> Right, so there's definitely something messing with the outgoing
> traffic. I have no idea what could cause netperf to only be able to send
> traffic in one direction (perhaps Rick knows?). Do you have any sort of
> firewall between you and the internet that might interfere with traffic?
>
> -Toke
>

[-- Attachment #2: Type: text/html, Size: 7820 bytes --]

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1.1: Type: text/plain, Size: 3855 bytes --]

yeah  with wired network it works:

1)  netperf-wrapper rrul netperf.bufferbloat.net

Test data is in ./rrul-2015-04-21T204050.322476.json.gz (use with -i to
format)

and output is result2.png(is attached).

2)  netperf-wrapper rrul -p all_scaled -l 60 -H netperf.bufferbloat.net

Graph(result3.png) is made.
But it does not tell anything at terminal(Like in 1st case Test data is
in....... or command never stops)
Seems like it keeps running.
Is it normal?


Also wanna know about what could be issue with my wifi?



On Tue, Apr 21, 2015 at 8:21 PM, sahil grover <sahilgrover013@gmail.com>
wrote:

> No,Firewall is disabled.
>
> tried with wired network
>
> 1. netperf -D -0.20 -H netperf.bufferbloat.net -t TCP_STREAM
>
> MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to
> atl.richb-hanover.com () port 0 AF_INET : demo
> Interim result:    0.40 10^6bits/s over 0.324 seconds ending at
> 1429627638.283
> Interim result:    0.28 10^6bits/s over 0.462 seconds ending at
> 1429627638.745
> Interim result:    1.61 10^6bits/s over 0.244 seconds ending at
> 1429627638.989
> Interim result:    0.98 10^6bits/s over 0.401 seconds ending at
> 1429627639.390
> Interim result:    0.51 10^6bits/s over 1.034 seconds ending at
> 1429627640.424
> Interim result:    0.44 10^6bits/s over 1.491 seconds ending at
> 1429627641.914
> Interim result:    0.27 10^6bits/s over 2.418 seconds ending at
> 1429627644.333
> Interim result:    0.70 10^6bits/s over 1.308 seconds ending at
> 1429627645.641
> Interim result:    0.49 10^6bits/s over 1.334 seconds ending at
> 1429627646.975
> Interim result:    0.40 10^6bits/s over 0.984 seconds ending at
> 1429627647.959
> Recv   Send    Send
> Socket Socket  Message  Elapsed
> Size   Size    Size     Time     Throughput
> bytes  bytes   bytes    secs.    10^6bits/sec
>
> 174760  16384  16384    13.07       0.38
>
>
>
>
> 2.  netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM
>
>
> MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to kau.toke.dk ()
> port 0 AF_INET : demo
> Interim result:    0.53 10^6bits/s over 0.247 seconds ending at
> 1429627754.173
> Interim result:    0.30 10^6bits/s over 0.437 seconds ending at
> 1429627754.610
> Interim result:    1.70 10^6bits/s over 0.462 seconds ending at
> 1429627755.072
> Interim result:    1.24 10^6bits/s over 0.424 seconds ending at
> 1429627755.496
> Interim result:    0.48 10^6bits/s over 1.378 seconds ending at
> 1429627756.874
> Interim result:    0.39 10^6bits/s over 2.686 seconds ending at
> 1429627759.560
> Interim result:    0.47 10^6bits/s over 2.215 seconds ending at
> 1429627761.775
> Interim result:    0.43 10^6bits/s over 2.151 seconds ending at
> 1429627763.926
> Recv   Send    Send
> Socket Socket  Message  Elapsed
> Size   Size    Size     Time     Throughput
> bytes  bytes   bytes    secs.    10^6bits/sec
>
>  87380  16384  16384    12.81       0.42
>
>
>
>
>
> On Tue, Apr 21, 2015 at 7:00 PM, Toke Høiland-Jørgensen <toke@toke.dk>
> wrote:
>
>> sahil grover <sahilgrover013@gmail.com> writes:
>>
>> > netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM
>> >
>> > Shows:
>> >
>> > MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to kau.toke.dk
>> () port 0
>> > AF_INET : demo
>> > Recv Send Send
>> > Socket Socket Message Elapsed
>> > Size Size Size Time Throughput
>> > bytes bytes bytes secs. 10^6bits/sec
>> >
>> > 87380 16384 16384 14.00 0.00
>>
>> Right, so there's definitely something messing with the outgoing
>> traffic. I have no idea what could cause netperf to only be able to send
>> traffic in one direction (perhaps Rick knows?). Do you have any sort of
>> firewall between you and the internet that might interfere with traffic?
>>
>> -Toke
>>
>
>

[-- Attachment #1.2: Type: text/html, Size: 8928 bytes --]

[-- Attachment #2: result2.png --]
[-- Type: image/png, Size: 123420 bytes --]

[-- Attachment #3: result3.png --]
[-- Type: image/png, Size: 117463 bytes --]

Re: [Codel] RRUL Test

[Rick Jones]

On 04/21/2015 06:30 AM, Toke Høiland-Jørgensen wrote:
> Right, so there's definitely something messing with the outgoing
> traffic. I have no idea what could cause netperf to only be able to send
> traffic in one direction (perhaps Rick knows?). Do you have any sort of
> firewall between you and the internet that might interfere with traffic?

Firewall is the first thing that comes to mind.  Second would be some 
sort of PathMTU discovery black-hole.

One way to check for that would be to run a TCP_RR test and see if that 
"works."  If it gets a non-zero result, it suggests an issue with large 
segments (the netperf default request and response size being one 
byte...).  If it gets a zero result, it suggests something else like 
perhaps firewall.

rick jones

Re: [Codel] RRUL Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 4775 bytes --]

Toke Sir,
 Thank you so much.
without your help it would never ever have been possible .
and i would have to spent many more days in installation itself only.
So thanks for everything.

Actually there is one more request.

kindly tell me something( more than something if possible) about its
results(result2.png,result3.png).

(1) what is it  showing in general?

(2) what can i interpret about these graphs related with bufferbloat.

i mean where is bufferbloat?

(3) Last is , in your articles,you have shown results with fq_codel and
codel using netperf-wrapper.
How you did that?
How can i  experiment with them(codel,fq_codel)?
what are the requirements?







On Tue, Apr 21, 2015 at 9:05 PM, sahil grover <sahilgrover013@gmail.com>
wrote:

> yeah  with wired network it works:
>
> 1)  netperf-wrapper rrul netperf.bufferbloat.net
>
> Test data is in ./rrul-2015-04-21T204050.322476.json.gz (use with -i to
> format)
>
> and output is result2.png(is attached).
>
> 2)  netperf-wrapper rrul -p all_scaled -l 60 -H netperf.bufferbloat.net
>
> Graph(result3.png) is made.
> But it does not tell anything at terminal(Like in 1st case Test data is
> in....... or command never stops)
> Seems like it keeps running.
> Is it normal?
>
>
> Also wanna know about what could be issue with my wifi?
>
>
>
> On Tue, Apr 21, 2015 at 8:21 PM, sahil grover <sahilgrover013@gmail.com>
> wrote:
>
>> No,Firewall is disabled.
>>
>> tried with wired network
>>
>> 1. netperf -D -0.20 -H netperf.bufferbloat.net -t TCP_STREAM
>>
>> MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to
>> atl.richb-hanover.com () port 0 AF_INET : demo
>> Interim result:    0.40 10^6bits/s over 0.324 seconds ending at
>> 1429627638.283
>> Interim result:    0.28 10^6bits/s over 0.462 seconds ending at
>> 1429627638.745
>> Interim result:    1.61 10^6bits/s over 0.244 seconds ending at
>> 1429627638.989
>> Interim result:    0.98 10^6bits/s over 0.401 seconds ending at
>> 1429627639.390
>> Interim result:    0.51 10^6bits/s over 1.034 seconds ending at
>> 1429627640.424
>> Interim result:    0.44 10^6bits/s over 1.491 seconds ending at
>> 1429627641.914
>> Interim result:    0.27 10^6bits/s over 2.418 seconds ending at
>> 1429627644.333
>> Interim result:    0.70 10^6bits/s over 1.308 seconds ending at
>> 1429627645.641
>> Interim result:    0.49 10^6bits/s over 1.334 seconds ending at
>> 1429627646.975
>> Interim result:    0.40 10^6bits/s over 0.984 seconds ending at
>> 1429627647.959
>> Recv   Send    Send
>> Socket Socket  Message  Elapsed
>> Size   Size    Size     Time     Throughput
>> bytes  bytes   bytes    secs.    10^6bits/sec
>>
>> 174760  16384  16384    13.07       0.38
>>
>>
>>
>>
>> 2.  netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM
>>
>>
>> MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to kau.toke.dk
>> () port 0 AF_INET : demo
>> Interim result:    0.53 10^6bits/s over 0.247 seconds ending at
>> 1429627754.173
>> Interim result:    0.30 10^6bits/s over 0.437 seconds ending at
>> 1429627754.610
>> Interim result:    1.70 10^6bits/s over 0.462 seconds ending at
>> 1429627755.072
>> Interim result:    1.24 10^6bits/s over 0.424 seconds ending at
>> 1429627755.496
>> Interim result:    0.48 10^6bits/s over 1.378 seconds ending at
>> 1429627756.874
>> Interim result:    0.39 10^6bits/s over 2.686 seconds ending at
>> 1429627759.560
>> Interim result:    0.47 10^6bits/s over 2.215 seconds ending at
>> 1429627761.775
>> Interim result:    0.43 10^6bits/s over 2.151 seconds ending at
>> 1429627763.926
>> Recv   Send    Send
>> Socket Socket  Message  Elapsed
>> Size   Size    Size     Time     Throughput
>> bytes  bytes   bytes    secs.    10^6bits/sec
>>
>>  87380  16384  16384    12.81       0.42
>>
>>
>>
>>
>>
>> On Tue, Apr 21, 2015 at 7:00 PM, Toke Høiland-Jørgensen <toke@toke.dk>
>> wrote:
>>
>>> sahil grover <sahilgrover013@gmail.com> writes:
>>>
>>> > netperf -D -0.20 -H netperf-eu.bufferbloat.net -t TCP_STREAM
>>> >
>>> > Shows:
>>> >
>>> > MIGRATED TCP STREAM TEST from 0.0.0.0 () port 0 AF_INET to kau.toke.dk
>>> () port 0
>>> > AF_INET : demo
>>> > Recv Send Send
>>> > Socket Socket Message Elapsed
>>> > Size Size Size Time Throughput
>>> > bytes bytes bytes secs. 10^6bits/sec
>>> >
>>> > 87380 16384 16384 14.00 0.00
>>>
>>> Right, so there's definitely something messing with the outgoing
>>> traffic. I have no idea what could cause netperf to only be able to send
>>> traffic in one direction (perhaps Rick knows?). Do you have any sort of
>>> firewall between you and the internet that might interfere with traffic?
>>>
>>> -Toke
>>>
>>
>>
>

[-- Attachment #2: Type: text/html, Size: 10303 bytes --]

Re: [Codel] RRUL Test

[Jonathan Morton]

> On 21 Apr, 2015, at 19:19, sahil grover <sahilgrover013@gmail.com> wrote:
> 
> Actually there is one more request.
> 
> kindly tell me something( more than something if possible) about its results(result2.png,result3.png).
> 
> (1) what is it  showing in general?
> 
> (2) what can i interpret about these graphs related with bufferbloat.
> 
> i mean where is bufferbloat?

Sahil - *please* try to develop and engage some critical-thinking and troubleshooting skills.  They will help you to answer questions like this without relying so heavily on others.  We can’t hold your hands every step of the way; we have work to do.

To guide you in the right direction, let me mention that in RRUL, the heavy traffic starts at the 5-second mark.  Before that, there is only probe traffic.  Now, armed with that knowledge, look at the latency graph and tell us what you see.

 - Jonathan Morton