[cdi-dev] [JBoss JIRA] (CDI-570) Weld shutdown in Java SE during asynch event calling

[Mark Struberg (JIRA)]

    [ https://issues.jboss.org/browse/CDI-570?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13127598#comment-13127598 ] 

Mark Struberg commented on CDI-570:
-----------------------------------

Yes, it's a whole category of problems. This not only affects shutdown but also normal runtime. If a synchronous event gets stuck then you can easily detect it. But if an async event get stuck then it might remain unnoticed for quite some time. Deadlocks might also occur much easier.

We might probably think about introducing something like @AbortAfter(60000) ?
Of course it makes all the CDI impls more heavyweight once we start with such things.

> Weld shutdown in Java SE during asynch event calling
> ----------------------------------------------------
>
>                 Key: CDI-570
>                 URL: https://issues.jboss.org/browse/CDI-570
>             Project: CDI Specification Issues
>          Issue Type: Feature Request
>          Components: Events
>    Affects Versions: 2.0-EDR1
>            Reporter: Michael Remijan
>
> In a Java SE environment, there is no way to tell if there are async observers are in the process of being called when the time comes for the JVM to shut down.  My assumption is the JVM should wait until all the async observers are complete before shutting down.  Or perhaps this behavior can be configured through the CDI object, system property, or beans.xml 



--
This message was sent by Atlassian JIRA
(v6.4.11#64026)