users@jersey.java.net

Re: [Jersey] Comment regarding http-thread-pool Idle Thread Timeout setting

From: Oleksiy Stashok <Oleksiy.Stashok_at_Sun.COM>
Date: Tue, 03 Aug 2010 11:30:59 +0200

Hi Noah,

which Glassfish version are you using? I remember we had naming issues
with several timeout values. But it had to be fixed.

Thanks.

WBR,
Alexey.

On Aug 3, 2010, at 11:25 , Paul Sandoz wrote:

> Hi Noah,
>
> Also the CC users_at_glassfish.dev.java.net as you may get more
> responses on that list.
>
> Paul.
>
> On Aug 2, 2010, at 11:32 PM, NBW wrote:
>
>> Hi folks,
>>
>> I have a question about the behavior I am seeing with the http-
>> thread-pool Idle Thread Timeout value. I have a Jersey resource
>> endpoint with receives a call. The resource class looks something
>> like this:
>>
>> @Path("/myResource")
>> @RequestScoped
>> public class MyResource {
>>
>> @EJB
>> MyEJB myEJB;
>>
>> @POST
>> @Consumes("application/x-www-form-urlencoded")
>> @Produces("text/plain")
>> public String doSomethingLongRunning(@FormParam("name") final
>> String name) throws DoSomethingLongRunningExeception {
>>
>> does some long running task
>>
>> return "ok";
>>
>> }
>> }
>>
>> So that's the gist of what if going on. Now what I am seeing is
>> that if the long running task takes longer then the value of http-
>> thread-pool Idle Thread Timeout, the thread running this task (an
>> http-thread-pool thread) will receive an InterruptedException.
>>
>> If this is the intended behavior then calling this value "Idle
>> Thread Timeout" is a bit misleading because the thread being
>> interrupted need not be "idle" for this to happen it just needs to
>> be RUNNING for longer then this timeout value.
>>
>> -Noah
>>
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe_at_glassfish.dev.java.net
> For additional commands, e-mail: users-help_at_glassfish.dev.java.net
>