Hi Reto, all

I think I found the reason for that. See STANBOL-651 [1] for details.
I am currently in the progress of fixing this.

best
Rupert

[1] https://issues.apache.org/jira/browse/STANBOL-651

On Fri, Mar 16, 2012 at 11:21 AM, Rupert Westenthaler
<[email protected]> wrote:
>
> On 16.03.2012, at 10:47, Reto Bachmann-Gmür wrote:
>
>> Hello,
>>
>> I'm having issues with jobs that bever result as finished and
>> EventJobManagerImpl.enhanceContent thus never ends waiting.
>>
>> I think it would be good to have a timeout in EventJobManagerImpl for it to
>> stop waiting for jobs after a while (and kiling the job if possible).
>>
>> I'm not yet sure about the reason for my Job never resulting as finished,
>> but a potential cause I think is the exception handling in
>> EnhancementJobHanlder: EnigneExceptions are caught and cause the job to be
>> marked as failed, while other exceptions don't result in the job being
>> marked as failed.
>>
>> Any thoughts or experiences on failing jobs?
>>
>
> While developing the EventJobManager this was indeed a problem.  My solution 
> was than to add a try catch(Throwable t) covering the event processing in the 
> EnhancementJobHandler#handleEvent(..) that sets the Job as failed within the 
> catch clause.
>
> This solved the problem with dealing of RuntimeException thrown by 
> EnhancementEngines.
>
> So I assume that the problem you encounter is
>
> (1) an engine that does not finish
> (2) an deadlock
> (3) an runtime exception within the EventJobManager implementation
>
> Reto can you reproduce this? If yes a DEBUG level log would really help to 
> trace this down
>
> regarding
>
>> I think it would be good to have a timeout in EventJobManagerImpl for it to
>> stop waiting for jobs after a while (and kiling the job if possible).
>
> That would need further investigation.
>
> The felix implementation of the EventAdmin [1] provides the possibility to 
> configure a Timeout ("org.apache.felix.eventadmin.Timeout").
> However if this happens the EventHandler is blacklisted. This would cause 
> that all event handling done by the EventJobManager would be stopped as all 
> Events are precessed by the same Handler. Because of that I was not able to 
> use this and had to deactivate Timeouts for the used Topic.
>
> If someone knows an other possibility to implement this I would be clearly 
> interested
>
> best
> Rupert
>
>
>
> [1] http://felix.apache.org/site/apache-felix-event-admin.html
>
>
>
> best
> Rupert
>
>> Cheers,
>> Reto
>



-- 
| Rupert Westenthaler             [email protected]
| Bodenlehenstraße 11                             ++43-699-11108907
| A-5500 Bischofshofen

Reply via email to