Hi Sebastian,

Following your recommandations, I tried with several step-size values, with no luck.
It does not seem to help the TCP upload graph.

I do not want to bother ML users with the graphs, I copy them to Dropbox:

https://www.dropbox.com/s/qxrdxvtqs7d6hfp/stepsize0.1.png?dl=0
https://www.dropbox.com/s/j54gaofor3ktjyt/stepsize0.5.png?dl=0
https://www.dropbox.com/s/i5fdqwldabx0820/stepsize0.05.png?dl=0
https://www.dropbox.com/s/dxf0ihlghy50d8b/stepsize1.png?dl=0

Thanks for your help,
Manu.


On Mon, Jul 31, 2017 at 8:13 PM, Sebastian Moeller <moeller0@gmx.de> wrote:
Hi,

I believe you need to set the step size to something bigger:
from "run-flent --help"
-s STEP_SIZE, --step-size STEP_SIZE
                        Measurement data point step size.

For a slow link there are simply no data from netperf for a given "sampling" point and hence you get the intermittent plots. As far as I know there is unfortunately no way to specify the step size independent for both directions, so for non-choppy plots in your slow direction you will also get less time resolution in your fast direction...

Hope that helps




On Jul 31, 2017, at 19:36, Emmanuel Blot <eblot.ml@gmail.com> wrote:

Hi all,

I’m a new flent user.

I’m trying to check my SQM settings on OpenWRT against bufferbloat.

For some reason, the TCP download and ping graphs for RRUL sessions
give sensible results, whereas the TCP upload is broken into very
small segments (even closer to dots actually), see the attached
sample.

What am I doing wrong ?

command line: flent rrul -p all_scaled -l 60 -H host -o output.png

$ flent --version
Flent v1.0.1.
Running on Python 3.5.2 (default, Oct 11 2016, 04:59:56)  [GCC 4.2.1
Compatible Apple LLVM 8.0.0 (clang-800.0.38)].
Using matplotlib version 2.0.1 on numpy 1.12.1.
Using PyQt5 version 5.9.


Thanks,
Manu.
_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org