[ 
https://jira.codehaus.org/browse/MSHARED-295?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=335546#comment-335546
 ] 

Frank Jakop edited comment on MSHARED-295 at 11/11/13 3:32 AM:
---------------------------------------------------------------

We ran in a similar problem situation with surefire 2.9 / plexus-utils 1.5.1. 
Our CI server did not kill running java processes on test timeout.
We upgraded to latest surefire (2.16) and saw cli classes relocated to 
maven-shared-utils.
Since we're not spawning further processes from our test JVM we can propose a 
very simple patch wich solves the problem for our needs.
Perhaps this is more likely to be accepted until Andrey's patch is fully 
working. (see MSHARED-295-no-cmd.patch)

                
      was (Author: fj):
    We ran in a similar problem situation with surefire 2.9 / plexus-utils 
1.5.1. Our CI server did not kill running java processes on test timeout.
Since we're not spawning further processes from our test JVM we can propose a 
very simple patch wich solves the problem for our needs.
Perhaps this is more likely to be accepted until Andrey's patch is fully 
working. (see MSHARED-295-no-cmd.patch)

                  
> Non reliable killing of processes by CommandLineUtils
> -----------------------------------------------------
>
>                 Key: MSHARED-295
>                 URL: https://jira.codehaus.org/browse/MSHARED-295
>             Project: Maven Shared Components
>          Issue Type: Improvement
>          Components: maven-shared-utils
>            Reporter: Andrey Klochkov
>            Assignee: Kristian Rosenvold
>         Attachments: MSHARED-295-no-cmd.patch, MSHARED-295.patch
>
>
> CommandLineUtils is used in Maven-Surefire to start forks which execute 
> tests. It is a well known issue that sometimes child processes are not killed 
> correctly. This is a known limitation of JVM, and the only reliable way to 
> implement it would be platform specific.
> Bug report in Surefire JIRA:
> http://jira.codehaus.org/browse/SUREFIRE-773
> JVM bug report. It's Windows specific, but the problem exists on Linux and 
> OSX as well.
> http://bugs.sun.com/view_bug.do?bug_id=4770092
> I'm proposing either 1) to implement several platform specific 
> implementations of ProcessHook or 2) to make the mechanism of killing 
> processes extendable for clients of the library.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Reply via email to