Hi Noah,
> I confirmed that Request Timeout is broken in 3.1.1 but it appears
> fixed in
> 3.1.2 so I won't file a bug for that.
Ok.
> I tried testing the Keep-Alive timeout w/3.1.2 by dialing it down to
> 5s but
> the endpoint ran for the full 25s and never timed out. Before I file a
> bug on
> this I just wanted to confirm this was a valid test for this timeout.
not really. You can test keep-alive timeout w/ telnet.
Telnet say to localhost 8080, make HTTP request
------------------------------------------
GET / HTTP/1.1
Host: localhost:8080
------------------------------------------
You'll see the response from server, then do nothing, just follow
console... telnet should report that connection is broken in 5 seconds
(if you set keep-alive timeout to 5 sec)
> Lastly, I turned down the http-thread-pool "Idle thread timeout"
> property to
> 5s and hit my endpoint which did not timeout, however, based on your
> description of this timeout it sounds like this probably isn't a valid
> test
> for this property. If that's the case and it's not do you have any
> thoughts
> on an approach to do this? Is there a test case SVN perhaps?
To test this timeout - you'll need to enable GF threads monitoring,
configure thread pool, so its min and max value are different, say 5 and 10.
Then you can use your timeout servlet and make say 11 parallel requests
to this servlet - you should see GF created 10 threads... after all
requests got processed, wait for "idle thread timeout" seconds and check
the number threads in a thread-pool... it should come back to 5.
Hope this will help.
Thanks.
WBR,
Alexey.
>
> Thanks,
>
> -Noah
>
>
> --
>
> [Message sent by forum member 'emailnbw']
>
> View Post: http://forums.java.net/node/883907
>
>