[Bloat] Realtime Response Under Load (RRUL) Test

[Bloat] Realtime Response Under Load (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: 544 bytes --]

Re: [Bloat] Realtime Response Under Load (RRUL) Test

[David Lang]

[-- Attachment #1: Type: TEXT/Plain, Size: 1198 bytes --]

On Fri, 17 Apr 2015, sahil grover wrote:

> 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?

whatever you are running

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

what are you trying to test? usually you don't get much value in testing a 
when you go through 
a router that is a bottleneck. so testing from one machine on your network to 
another machine on your network isn't going to find anything

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

It boils down to

1. do a latency test with no other load ( ping works )

2. start a large upload and see what happens to latency during the upload

3. start a large download and see what happens to latency during the download

4. start an upload and download and see what happens to latency.

the netperf test does this by having you run a netperf server on one system (on 
the other side of the router that is your bottleneck), then run the test from 
a machne on your network, and it does allt he steps above for you.

David Lang

[-- Attachment #2: Type: TEXT/PLAIN, Size: 140 bytes --]

_______________________________________________
Bloat mailing list
Bloat@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/bloat

Re: [Bloat] Realtime Response Under Load (RRUL) Test

[David Lang]

On Mon, 20 Apr 2015, sahil grover wrote:

> David Sir,
>
> Thanks for answering.
>
> can you please tell me how to interpret its output?
>
> By   "netperf-wrapper rrul netperf.bufferbloat.net" ( i chose netserver
> that is running on netperf.bufferbloat.net)
> output(result.png) is attached.
>
> please tell me first what does it say generally?
> secondly,  anything related with bufferbloat point of view.

latency should not increase significantly under load. The fact that it does 
indicates that you have a problem. The fact that the latency can increase to 
multi-second values for many people indicates a big problem.

take a look at the new test that's available at 
https://www.dslreports.com/speedtest and then look at the results tab and see 
how the latency increases during the download and upload tests.

If you have a system without excessive buffering, you would not see massive 
increases in latency.

Contact me off-list if you want a deeper discussion of the history of 
bufferbloat and what's going on to try and solve it.

David Lang

Re: [Bloat] Realtime Response Under Load (RRUL) Test

[sahil grover]

[-- Attachment #1: Type: text/plain, Size: 1487 bytes --]

>Contact me off-list if you want a deeper discussion of the history of
bufferbloat and what's going on to try and solve it.

Sure, i will. You made my day  :)
because by reading papers/articles, i didn't get that i gained from you
people.
Infact always wanna hear something like this.

So Thanks a lot for being so kind.


On Mon, Apr 20, 2015 at 9:15 PM, David Lang <david@lang.hm> wrote:

> On Mon, 20 Apr 2015, sahil grover wrote:
>
>  David Sir,
>>
>> Thanks for answering.
>>
>> can you please tell me how to interpret its output?
>>
>> By   "netperf-wrapper rrul netperf.bufferbloat.net" ( i chose netserver
>> that is running on netperf.bufferbloat.net)
>> output(result.png) is attached.
>>
>> please tell me first what does it say generally?
>> secondly,  anything related with bufferbloat point of view.
>>
>
> latency should not increase significantly under load. The fact that it
> does indicates that you have a problem. The fact that the latency can
> increase to multi-second values for many people indicates a big problem.
>
> take a look at the new test that's available at
> https://www.dslreports.com/speedtest and then look at the results tab and
> see how the latency increases during the download and upload tests.
>
> If you have a system without excessive buffering, you would not see
> massive increases in latency.
>
> Contact me off-list if you want a deeper discussion of the history of
> bufferbloat and what's going on to try and solve it.
>
> David Lang
>

[-- Attachment #2: Type: text/html, Size: 2336 bytes --]