You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@mesos.apache.org by "Alexander Rukletsov (JIRA)" <ji...@apache.org> on 2016/10/13 21:16:20 UTC

[jira] [Commented] (MESOS-6387) Improve reporting of parallel test runner.

    [ https://issues.apache.org/jira/browse/MESOS-6387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15573204#comment-15573204 ] 

Alexander Rukletsov commented on MESOS-6387:
--------------------------------------------

The issue I have with parallel tests is that it is unclear how many tests have failed. I have to scroll up to gather a list of failed tests, because there is no aggregation among shards. For example, in this run two tests have actually failed, which is unclear from the output:
{noformat}
[  FAILED  ] 1 test, listed below:
[  FAILED  ] DefaultExecutorTest.KillTaskGroupOnTaskFailure

 1 FAILED TEST
  YOU HAVE 4 DISABLED TESTS



[FAIL]
make[3]: *** [check-local] Error 2
make[2]: *** [check-am] Error 2
make[1]: *** [check] Error 2
make: *** [check-recursive] Error 1
{noformat}

> Improve reporting of parallel test runner.
> ------------------------------------------
>
>                 Key: MESOS-6387
>                 URL: https://issues.apache.org/jira/browse/MESOS-6387
>             Project: Mesos
>          Issue Type: Improvement
>          Components: test
>            Reporter: Benjamin Bannier
>              Labels: mesosphere, newbie, python
>
> We should improve the logging of the parallel test runner. Improved logging seems to require some parsing of GoogleTest output, e.g., to prevent interleaving test output from concurrently executing tests.
> We should add a verbose mode which can print results as they arrive from shards in order to not let users wonder if tests are executing.
> We should also provide a way to properly unify output from different shards, e.g., report passed and failed tests in a single list instead of listing all shards separately (this e.g., makes tests from failed shards reported first harder to discover).
> Distinguishing reports from tests run in parallel and sequentially might be useful as well.



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