No problem. Thanks for your kind words.

To be honest, that’s a really complex question, and we don’t have space here to go into all the details.

I would suggest you start with this series of articles: http://www.ibm.com/developerworks/websphere/techjournal/0604_singh/0604_singh.html. They are for WebSphere Application Server, so some of the details will be different. Specifically, when you’re working with WPS/WESB, you tend to configure design-time security through the Deployment Editor on the Process Server project in question, rather than at the ‘lower-level’ J2EE configuration layers the article is talking about. The panels are very similar, however.

Its been a long time since we interacted. Hope things are well at you end.

This time around I have a couple of questions around WS-Security.

First a brief about the environment:
WID 6.0.1
WPS 6.0.1.6
ESB 6.0.1.6
TAM 5.1

What we are trying to do is secure all our web services using Basic Authentication, LTPA and Username token where either one of LTPA or Username token can be used.

My first question is whether there is any how-to article/document around what has to be done in the WID projects to configure basic auth and either LTPA or Username token.

Before I jump into my second question, a little background on how the services are layered in this context. All mediation modules are on ESB and all BPEL ones are (obviously) on WPS. WPS services generally invoke ESB services which integrate with various back-ends. This invocation is using the WS binding. Moreover, though the WPS services are exposed as Web Services, they are not to be invoked directly from outside. There is an ESB Facade deployed which acts as the gateway for the world outside. Between the ESB Facade and the WPS module, the invocation is via WS binding again. Hence, the layers look like as follows:
ESB Facade -> BPEL -> ESB

All the services in the layer will have Basic Auth, LTPA and Username token enabled on them.

The question is if the Facade receives a set of user credentials in the HTTP Request Header and either a LTPA or Username token from the consumer, can the same be propagated to all the other layers? If so, how can this be configured?

Thanking you in advance for your time. Please feel free to get back to me for any further clarifications.

Regards,

I have a scenario in ESB, where currently i have configured three queues(Queue1,Queue2,Queue3)
based on my routing logic the message will be routed to the respective queues (Queue1,Queue2,Queue3) .

In future we might add more queues to the same routing logic, i.e for a single condition the message can be routed to more than one queues. This should be done without modifying the Mediation Module.

The goal is to make the mediation process configurable, so that it can be used to route and filter incoming messages to multiple queues to help scale the message processing.

My version of WID is 6.1.2

Thanks a lot, it worked very well.

Also, is there any disadvantage of using Foreign bus, at what scenario do we need to use this foreign bus

when using foreign bus, does both the server should be be in clustered.

If not in cluster how to do?

I think you will want to take a look at using a foreign bus:

http://publib.boulder.ibm.com/infocenter/wasinfo/v6r0/index.jsp?topic=/com.ibm.websphere.pmc.nd.doc/tasks/tjj0074_.html

Hope that helps,
Andrew.