You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tinkerpop.apache.org by Stephen Mallette <sp...@gmail.com> on 2016/04/18 12:30:12 UTC
running benchmarks
Ted, I went to run benchmarks this morning with maven. The instructions
mentioned that you needed to do this:
mvn clean test -DskipBenchmarks=false -DdefaultForks=5
-DmeasureIterations=20 -DwarmupIterations=20
The "test" target generated errors for me which had me confused until I
recalled that our "test" target sorta doesn't work so well in any case
because of shading. I ended up running this with:
mvn clean test -pl gremlin-benchmark -DskipBenchmarks=false
-DdefaultForks=5 -DmeasureIterations=20 -DwarmupIterations=20
which ran to success. Should the documentation be changed to that command
or am I missing something in how you run the benchmarks from maven?
Re: running benchmarks
Posted by Stephen Mallette <sp...@gmail.com>.
it might work that way if you cd into gremlin-benchmark first but i prefer
to just use -pl gremlin-benchmark option.
On Mon, Apr 18, 2016 at 10:24 AM, Ted Wilmes <tw...@gmail.com> wrote:
> Looks like the documentation should be changed. I'm not sure what voodoo I
> did to make vanilla `mvn clean test` work because sure enough, it blows up
> on my end too. Regardless, I'll get the docs updated. Thanks for catching
> this.
>
> --Ted
>
> On Mon, Apr 18, 2016 at 3:30 AM, Stephen Mallette <sp...@gmail.com>
> wrote:
>
> > Ted, I went to run benchmarks this morning with maven. The instructions
> > mentioned that you needed to do this:
> >
> > mvn clean test -DskipBenchmarks=false -DdefaultForks=5
> > -DmeasureIterations=20 -DwarmupIterations=20
> >
> > The "test" target generated errors for me which had me confused until I
> > recalled that our "test" target sorta doesn't work so well in any case
> > because of shading. I ended up running this with:
> >
> > mvn clean test -pl gremlin-benchmark -DskipBenchmarks=false
> > -DdefaultForks=5 -DmeasureIterations=20 -DwarmupIterations=20
> >
> > which ran to success. Should the documentation be changed to that
> command
> > or am I missing something in how you run the benchmarks from maven?
> >
>
Re: running benchmarks
Posted by Ted Wilmes <tw...@gmail.com>.
Looks like the documentation should be changed. I'm not sure what voodoo I
did to make vanilla `mvn clean test` work because sure enough, it blows up
on my end too. Regardless, I'll get the docs updated. Thanks for catching
this.
--Ted
On Mon, Apr 18, 2016 at 3:30 AM, Stephen Mallette <sp...@gmail.com>
wrote:
> Ted, I went to run benchmarks this morning with maven. The instructions
> mentioned that you needed to do this:
>
> mvn clean test -DskipBenchmarks=false -DdefaultForks=5
> -DmeasureIterations=20 -DwarmupIterations=20
>
> The "test" target generated errors for me which had me confused until I
> recalled that our "test" target sorta doesn't work so well in any case
> because of shading. I ended up running this with:
>
> mvn clean test -pl gremlin-benchmark -DskipBenchmarks=false
> -DdefaultForks=5 -DmeasureIterations=20 -DwarmupIterations=20
>
> which ran to success. Should the documentation be changed to that command
> or am I missing something in how you run the benchmarks from maven?
>