You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@struts.apache.org by "Falko Modler (Commented) (JIRA)" <ji...@apache.org> on 2011/10/13 00:41:11 UTC

[jira] [Commented] (WW-3691) BackgroundProcess should use a java.util.concurrent.Executor alternatively to spawning a new thread

    [ https://issues.apache.org/jira/browse/WW-3691?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13126216#comment-13126216 ] 

Falko Modler commented on WW-3691:
----------------------------------

Thanks for the quick reply!

I don't think this a sufficient solution because in your patch you just delegate the execution of the Runnable/Thread to the interceptor which now implements Executor. But because there is no Thread.start() anymore, there is no more "background processing thread" anymore.

My idea is more like:
Inject an Executor into the interceptor (@Inject(required=false)) which is then passed to the BackgroundProcess constructor.

I will try it out myself this weekend and I will provide a patch if everything works as expected.
                
> BackgroundProcess should use a java.util.concurrent.Executor alternatively to spawning a new thread
> ---------------------------------------------------------------------------------------------------
>
>                 Key: WW-3691
>                 URL: https://issues.apache.org/jira/browse/WW-3691
>             Project: Struts 2
>          Issue Type: Improvement
>          Components: Core Interceptors
>    Affects Versions: 2.2.3.1
>            Reporter: Falko Modler
>         Attachments: WW-3691.patch
>
>
> Every new instance of org.apache.struts2.interceptor.BackgroundProcess spawns a new Thread (see constructor), no thread pooling is used.
> Besides problems in environments where some container might need to manage the creation of threads, this issue also prevents certain performance/efficiency optimizations via ThreadLocal from taking full effect. E.g.: We use ThreadLocals for Random and SimpleDateFormat and those thread local instances "are lost" when a new Thread is created.
> Therefore BackgroundProcess should be given a new constructor that takes a java.util.concurrent.Executor instance that is used to execute the Runnable.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira