On Fri, Jan 21, 2011 at 8:57 AM, Deepal jayasinghe <[email protected]>wrote:
> On 1/20/2011 5:12 AM, Kanchana Welagedara wrote: > > Hi Folks; > > > I'm trying to develop a tool which to identify web service > dependencies,stubs/skeleton granularity , > > What kind of dependencies? are you referring to something like BPEL? > service /client based jar dependencies/any > > service mapping and impact change analysis tool. > I have some statics data (method call ,order of the method invocation) > generated based on some .Net client stubs which consumes a Axis java > service.There is a middle frame work where that end point URL,invocation > method in client stub can be different from the actual service URL and the > actual service method. Based these statistical data (not the dynamic > environment > )how would i recognized that service served by which client and vice verca. > > I think you need to get some data from the middle framework you have > mentioned. > Can't we locate the actual WS service ,method name from EPR ? how do we extract EPR value from actual service when comparing WSDL file ,if it's not a the service url given in WSDL? Isn't is proxy which should always know the actual URL of the service? Following using handler tutorial http://blogs.deepal.org/2009/02/writing-axis2-module.html forgetting about the middle tire dependencies in my system.If I set a handler in the axis 2 java soap engine will it invoke for all the services ? and get me a token from each client which reach all the deployed services and identify the each client call uniquely which consumes which service ? how can we store such Data ? in call ? XML? DB? Seems this is a approach to untouched the skeletons in server side.isnt it ?How common is this handler in any soap engine?Should be a generic isn't it? > > Deepal > > There are not any registry information which I can use to map these > services. > If I can get the meta data of the service and client implemetation level > what are the artifacts which I can use.(example in Axis java .WSDD.xml ?) > > Really appreciate you input in this > > Thanks > Kanchana > > >
