You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@s2graph.apache.org by "DOYUNG YOON (JIRA)" <ji...@apache.org> on 2016/08/29 13:05:21 UTC

[jira] [Created] (S2GRAPH-99) Reduce number of test run on test cases on org.apache.s2graph.rest.play.benchmark

DOYUNG YOON created S2GRAPH-99:
----------------------------------

             Summary: Reduce number of test run on test cases on org.apache.s2graph.rest.play.benchmark
                 Key: S2GRAPH-99
                 URL: https://issues.apache.org/jira/browse/S2GRAPH-99
             Project: S2Graph
          Issue Type: Sub-task
    Affects Versions: 0.1.0
            Reporter: DOYUNG YOON
            Assignee: DOYUNG YOON
             Fix For: 0.1.0


Currently SamplingBenchmarkSpec on s2rest_play project test different sampling methods on size 1000 edges averaging 500000 times. Since we are more interested on relative performance between different sampling method rather than absolute peformance, I am suggesting to reduce number of test run from 500000 to 10000(1/50) to make test run faster.

{noformat}
==================================================
Random Array Sampling: took 8696 ms
==================================================

==================================================
Shuffle Sampling: took 30263 ms
==================================================

==================================================
RNG Sampling: took 21161 ms
==================================================
[info] SamplingBenchmarkSpec
[info]
[info] sample should
[info] + sample benchmark
[info]
[info] Total for specification SamplingBenchmarkSpec
[info] Finished in 2 minutes 5 seconds, 499 ms
{noformat}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)