Ah Yes  I remember reading that thread now :). Thanks

On Tue, Jul 22, 2008 at 10:28 PM, David Blevins <[EMAIL PROTECTED]> wrote:
>
> On Jul 22, 2008, at 5:40 AM, Manu George wrote:
>
>> I had a look at this and am facing a certain behaviour that is
>> puzzling me. I create an EJB and from a client make a 1000 parallel
>> invocations. I set the poolSize as say 10 for the stateless container
>> but still all the 1000 invocations use only a single instance.
>> The ThreadPoolExecutor in ServicePool seems to be deciding how many
>> threads are executing in parallel and most of the time the execution
>> is happening serially. At much higher loads more instances are getting
>> created but that is decided by how many threads are being created by
>> the ThreadPoolExecutor. So the timeout variable which i worked back in
>> seems to be not to be getting used most of the time.
>
> It's actually getting decided by the client.  See this description from the
> "Remote client performance" thread.
>
> On Jul 13, 2008, at 5:41 PM, David Blevins wrote:
>
>> I also made it so the client can have at most one open connection per
>> server URL and all requests file through it.
>>
>> For the record you can actually get more connections per client by
>> slightly appending to your server url as so:
>>
>> ejbd://localhost:4201/a
>> ejbd://localhost:4201/b
>>
>> Everything after the "/" is currently ignored, but will allow you to get
>> another connection to the same server if you think you really need it.  I
>> actually saw pretty great performance with just one connection shared among
>> 10 client threads, but maybe you want a second connection open for very long
>> running requests as to not block up the shorter requests.
>
> I'd say the best way to test the pooling is to have a testcase that uses the
> LocalInitialContextFactory.  Then you can test as many threads as you need
> and have the container code more isolated than with including the
> client/server code as well.  Should also allow you to keep the test case in
> the openejb-core package.
>
> -David
>
>

Reply via email to