Hi Michele, In my case, there are different services running on axis2 . So if I set the thread pool size to 1 , i have a single thread serving every request (of anyone of my services) , doesn't it? If it is the case, this thread becames a bottleneck for the system and it is not a feasible solution.
Thanks a lot, Regards Francesco 2007/6/1, Michele Mazzucco <[EMAIL PROTECTED]>:
Paul, what about setting the axis2 thread pool size to 1?, this would automatically serialize all requests (I'm assuming that only that service is deployed). Michele On Fri, 2007-06-01 at 11:08 +0100, Paul Fremantle wrote: > Francesco > > Two questions! > > 1) Is this a one-way or a request reply operation. > 2) Have you considered using WSRM which provides inOrder delivery (in > other words, it gives each message a number and ensures they get > delivered in the order of the message numbers) > > Paul > > On 6/1/07, Francesco Leone <[EMAIL PROTECTED]> wrote: > > Hi, > > I have a problem with axis2 and its request management. The scenario is the > > sequent: there is a service deployed in application scope. > > This service make a blocking call to a synchronized queue. If I well > > understand Axis2 and its application scope, > > it creates a thread for every request for this service and every > > thread communicate with the same service object. So it is possible that many > > thread are blocked for the synchronized queue . The problem is that i want > > to service the request basing on its arrival time, It's important for my > > application semantic, so when the > > queue is free i want to notify the first thread arrived to axis2. > > Any suggestion about resolving this problem? > > Thanks, > > Regards > > Francesco Leone > > > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
