Emailed question: ODL with OpenStack - WCBench no result

[dfarrell07]

_Copying an email to this Issues so I can answer it publicly (docs for others, etc)._

I am running ODL Helium pre-built on Ubuntu. Integrated it with OpenStack 3 node architecture. All is working fine.

I installed WCBench in Ubuntu (made necessary changes in wcbench.sh file). The results are below:

:~/wcbench$ ./wcbench.sh -v -r
Collecting pre-test stats
Running CBench against ODL on 10.107.2.244:6653
cbench: controller benchmarking tool
running in mode 'latency'
connecting to controller at 10.107.2.244:6653
faking 16 switches offset 1 :: 10 tests each; 10000 ms per test
with 100000 unique source MACs per switch
learning destination mac addresses before the test
starting test with 0 ms delay after features_reply
ignoring first 1 "warmup" and last 0 "cooldown" loops
connection delay of 0ms per 1 switch(es)
debugging info is off
Collecting post-test stats
Collecting time-irrelevant stats
Average responses/second: 0.00

Why is the response zero? results.csv file is generated and attaching the same here.

stats.py results in follwoing error.

:~/wcbench$ ./stats.py -S
Traceback (most recent call last):
File "./stats.py", line 345, in
stats_mapstat
File "./stats.py", line 216, in compute_five_load_stats
self.compute_generic_stats("five_load", self.five_load_col)
File "./stats.py", line 249, in compute_generic_stats
generic_stats["min"] = int(numpy.amin(stats_col))
File "/usr/lib/python2.7/dist-packages/numpy/core/fromnumeric.py", line 2209, in amin
out=out, keepdims=keepdims)
File "/usr/lib/python2.7/dist-packages/numpy/core/_methods.py", line 21, in _amin
out=out, keepdims=keepdims)
ValueError: zero-size array to reduction operation minimum which has no identity

:~/wcbench$ ./stats.py -G
Traceback (most recent call last):
File "./stats.py", line 335, in
graph_map[graph](len%28graphs_to_build%29, graph_num+1)
File "./stats.py", line 228, in build_five_load_graph
"Five Minute Load", self.five_load_col)
File "./stats.py", line 277, in build_generic_graph
pyplot.plot(self.run_col, data_col, "go")
File "/usr/lib/pymodules/python2.7/matplotlib/pyplot.py", line 2987, in plot
ret = ax.plot(_args, *_kwargs)
File "/usr/lib/pymodules/python2.7/matplotlib/axes.py", line 4137, in plot
for line in self._get_lines(_args, *_kwargs):
File "/usr/lib/pymodules/python2.7/matplotlib/axes.py", line 317, in _grab_next_args
for seg in self._plot_args(remaining, kwargs):
File "/usr/lib/pymodules/python2.7/matplotlib/axes.py", line 295, in _plot_args
x, y = self._xy_from_xy(x, y)
File "/usr/lib/pymodules/python2.7/matplotlib/axes.py", line 237, in _xy_from_xy
raise ValueError("x and y must have same first dimension")
ValueError: x and y must have same first dimension

Is this because there is not input in the results.csv file? And is the output nill because the controller is busy with OpenStack?

[dfarrell07]

I installed WCBench in Ubuntu (made necessary changes in wcbench.sh file).

It'd be ideal if you'd contribute those back upstream. That's how Open Source projects stay healthy :)

Is this because there is not input in the results.csv file?

The stats.py helper isn't designed to handle results.csv files that are missing fields, or generally anything other than what's generated by a normal WCBench run. I raised #65 to track this feature request.

And is the output nill because the controller is busy with OpenStack?

I've seen 0.0 results a number of times, typically it's an ephemeral error that goes away after I restart the ODL process. However, since you've made a bunch of changes to ODL/WCBench to get it running on Ubuntu with OpenStack, I think it's fairly likely that part of that process is the issue. Would really need a ton more info. Could try just the Ubuntu changes, get working results merged upstream (we'll verify it works with some rigor at that point), then attempt ODL+OpenStack?

[Jaanki]

 It'd be ideal if you'd contribute those back upstream. That's how Open Source projects stay healthy :)

Ya sure. Added the repo here Is this correct way or should add it somewhere else? Am new to github and to contributing to open source projects :)

[dfarrell07]

Am new to github and to contributing to open source projects :)

Ah, well cool, welcome! :)

Added the repo here Is this correct way or should add it somewhere else?

No, that's not the right way (would be very confusing for users, "which project should I use?", much harder to maintain once they diverge).

As I describe very briefly in the README, you'll want to fork this repo (there's a button on the main page). That'll create a copy you own and can work with. Make committs to your fork to add your changes, then once they are ready submit a pull request (commonly called "PR", there will be a button on your fork's page) to merge your fork and its changes back into the main project. Once you submit it I'll be notified, review it, provide feedback and once it's ready I'll merge your commits into the master branch of WCBench.

There's tons of useful documentation about PRs, including:

https://gun.io/blog/how-to-github-fork-branch-and-pull-request/
http://stackoverflow.com/questions/14680711/how-to-do-a-github-pull-request
https://help.github.com/articles/using-pull-requests/

[dfarrell07]

Closing this issue, we can handle additional discussion in the PR.