[jira] [Created] (IGNITE-3814) Unstable results of compute benchmarks

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

[jira] [Created] (IGNITE-3814) Unstable results of compute benchmarks

Anton Vinogradov (Jira)
Ilya Suntsov created IGNITE-3814:
------------------------------------

             Summary: Unstable results of compute benchmarks
                 Key: IGNITE-3814
                 URL: https://issues.apache.org/jira/browse/IGNITE-3814
             Project: Ignite
          Issue Type: Bug
    Affects Versions: 1.7
         Environment: Ubuntu 14.04
jdk 1.7, jdk 1.8
            Reporter: Ilya Suntsov


I've run the following compute benchmarks:
* exec-compute
* run-compute
* broad-compute
* apply-compute
and got huge difference between the best and worst results e.g. for exec-compute we have the following numbers:
1 1222.38
2 1220.63
3 1044.53
4 1342.39
5 1393.57
5 1064.66

delta between 3-rd and 4-th runs is 28%.
I think we should re-work these benchmarks.



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