You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@camel.apache.org by "Claus Ibsen (JIRA)" <ji...@apache.org> on 2014/12/21 08:52:13 UTC

[jira] [Resolved] (CAMEL-8165) Async routing engine - Add insight into threads blocked waiting for callbacks

     [ https://issues.apache.org/jira/browse/CAMEL-8165?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Claus Ibsen resolved CAMEL-8165.
--------------------------------
    Resolution: Fixed

There is a new async processor await manager (yeah naming is hard) that handles this and allows insight and apis to force blocked threads to be freed, and as well during graceful shutdown of Camel itself.

https://cwiki.apache.org/confluence/display/CAMEL/AsyncProcessorAwaitManager

> Async routing engine - Add insight into threads blocked waiting for callbacks
> -----------------------------------------------------------------------------
>
>                 Key: CAMEL-8165
>                 URL: https://issues.apache.org/jira/browse/CAMEL-8165
>             Project: Camel
>          Issue Type: New Feature
>          Components: camel-core, jmx
>            Reporter: Claus Ibsen
>            Assignee: Claus Ibsen
>             Fix For: 2.15.0
>
>
> During routing there async routing engine may kick-in and depending on which components / EIPs are in use, and what options people have turned on/off, there is many combos.
> In some situations the routing engine may need to block a thread to force it be synchronous at a given point.
> We should add insight at runtime when these situations happens, so end users can monitor that. And also control the locks in case something bad happens. Or they are want to force "rouge" threads to terminate.
> These bad situations can be caused by 3rd party developed components, the JVM running out of memory, or java.lang.Error being thrown, or 3rd party takes long time to process, or never signal a callback, or whatever.
> This can cause threads in Camel to appears as stuck, and today to remedy this is to restart the JVM, or force a thread to terminate using some JVM tool.



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