[Battlemesh] battlemesh v6 test data postprocessing

Axel Neumann neumann at cgws.de
Tue May 20 11:52:50 UTC 2014


On 05/20/2014 07:56 AM, Henning Rogge wrote:
> Hi,
> 
> is there already a place where the raw data output of the experiments
> have been placed?


Guessing the quick answer to this: YES! OLSRv2 was working cool!


Some raw measurement data:
http://wibed.confine-project.eu/resultsdir/wbmv7-axn-16_2014-05-16_19-28-43
(stationary scenarios)

http://wibed.confine-project.eu/resultsdir/wbmv7-axn-17_2014-05-16_20-13-20
(broken crossed streams scenario)

http://wibed.confine-project.eu/resultsdir/wbmv7-axn-19_2014-05-16_21-35-33
(mobile scenarios)


You must choose the directory or tgz of a specific node to find the
relevant data according to this table:
https://github.com/axn/wbm2pdf/blob/master/wbmv7.tex#L188

Some few and preliminary results are here:
https://github.com/axn/wbm2pdf/blob/master/wbmv7.pdf?raw=true


/axel


> 
> Henning
> 
> On Tue, May 20, 2014 at 12:52 AM, Dave Taht <dave.taht at gmail.com> wrote:
>> On Mon, May 19, 2014 at 10:05 AM, Clauz <clauz at ninux.org> wrote:
>>> On 05/17/2014 11:41 AM, Clauz wrote:
>>>> On 05/16/2014 11:57 PM, Juliusz Chroboczek wrote:
>>>>>> During this battlemesh I would like to post-process the data from
>>>>>> Aalborg. I already produced some plots in the past days:
>>>>>
>>>>> Very strange results.  Are the pairs of nodes chosen the same for all
>>>>> the protocols?  Are the protocols running at the same time, or
>>>>> sequentially?
>>>>
>>>> The protocols were all running at the same time, but the tests were ran
>>>> sequentially. The pair of nodes was chosen randomly at each single test.
>>>> And, we didn't collect enough measurements to obtain the same set of
>>>> nodes for all the protocols.
>>>
>>> So, on Saturday I talked to Axel, who also did some postprocessing on
>>> WBMv6 data [*] and who asked me to double check.
>>>
>>> The script which has been used for all netperf tests is here:
>>> https://github.com/cl4u2/battlemeshv6_tests_postprocessing/blob/master/netperf_test2/scripts/netperf2.sh
>>>
>>> This can be also seen from the screenlogs, e.g.
>>> https://github.com/cl4u2/battlemeshv6_tests_postprocessing/blob/master/netperf_test2/data/screenlog.netperf
>>>
>>> So we had just one fixed source node and one fixed destination node all
>>> the time. All the protocols were running at the same time but the tests
>>> were ran in sequence.
>>>
>>> Sorry. This happens when you put your hands on stuff after one year,
>>> with bad documentation :(
>>>
>>> Juliusz, it would have been really super to have your advice and
>>> help during the last battlemesh events.
>>>
>>> I hope this year's postprocessing will be quicker and that the
>>> documentation will be better than WBMv6's.
>>
>> Did you give netperf-wrapper a shot this year?
>>
>>>
>>> Clauz
>>>
>>> [*] https://github.com/axn/wbm2pdf/blob/wbmv6/tex/wbmv6.pdf?raw=true
>>>
>>> _______________________________________________
>>> Battlemesh mailing list
>>> Battlemesh at ml.ninux.org
>>> http://ml.ninux.org/mailman/listinfo/battlemesh
>>
>>
>>
>> --
>> Dave Täht
>>
>> NSFW: https://w2.eff.org/Censorship/Internet_censorship_bills/russell_0296_indecent.article
>> _______________________________________________
>> Battlemesh mailing list
>> Battlemesh at ml.ninux.org
>> http://ml.ninux.org/mailman/listinfo/battlemesh
> _______________________________________________
> Battlemesh mailing list
> Battlemesh at ml.ninux.org
> http://ml.ninux.org/mailman/listinfo/battlemesh
> 



More information about the Battlemesh mailing list