| |
|
RSS Feed - WebSphere MQ Support
|
RSS Feed - Message Broker Support
|
| Accessing Services from ESB and Broker |
« View previous topic :: View next topic » |
| Author |
Message
|
| jbanoop |
 Posted: Tue Oct 03, 2006 10:36 pm Post subject: |
 |
|
Chevalier
Joined: 17 Sep 2005
Posts: 401
Location: SC
|
I personally have not used Websphere ESB but if you can put a message using JMS to a queue (in MQ) , that would be more than sufficent for the broker to pick it up and process using the flows.
Is there any other way to integrate between ESB and MB ..you are asking the wrong person..However webservices would definitely be another option.
It seems you are going round in circles.. wht exactly is the kind of communication you are expecting ..
Anoop |
|
| Back to top |
|
 |
| sajid08 |
| Posted: Thu Oct 05, 2006 12:00 am Post subject: |
|
|
Novice
Joined: 08 Sep 2006
Posts: 22
Location: Karachi
|
Actually we are in the design phase of the architecture we'll be using for integration, and mapping different IBM products on the architecture, so that we can decide which products should we buy, which would fulfil our needs.
Actually I didn't meant that Broker should pick up the message from the queue after ESB has put it there, Infact a service or application would get the message from there, was just wondering whether this is possible or not  .
The solution we are approaching is, we'd need to have queues on both ends, like, these would be the steps:
1. Application which requires a service, puts the message on the queue. (We want this and not the application directly talking to the broker or ESB so that the interface from applications to middleware remains same and if we change the middleware in the future it wont affect the applications interface to it bcz Applications are always talking to the queues and not to the middleware).
2. Broker or ESB picks up the message from there.
3. Transform it if needed, gets to know with some lookup that for which service it is meant, and put the message on that service's receive queue.
4. That service processes it and puts the message on the reply queue.
5. Broker or ESB picks that message from there, and puts that to another queue where from the calling application can pick it.
That means having two queues at each interface of the middleware, two at the Application requesting end and two at the service end.
What do you people think of this?
Regards,
Sajid. |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Thu Oct 05, 2006 2:55 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
| sajid08 wrote: |
That means having two queues at each interface of the middleware, two at the Application requesting end and two at the service end.
|
Not necessarily. For non persistent messages the application creates their own dynamic temporary queue for the reply message.
Multiple operations on the wsdl can/may use the same queue on the broker/requestor/provider...
The 2 queues at the service end may only apply to the Mainframe...
And then there is scalability which will multiply your number of queues by the number of qmgrs/brokers balancing the load....
Enjoy 
_________________
MQ & Broker admin |
|
| Back to top |
|
|
| sajid08 |
| Posted: Thu Oct 05, 2006 10:17 pm Post subject: |
|
|
Novice
Joined: 08 Sep 2006
Posts: 22
Location: Karachi
|
When you say
| Quote: |
| "For non persistent messages the application creates their own dynamic temporary queue for the reply message" |
1. does that mean applications manually creating queues, and how would that be achieved in web applications? For example, we have a web based Net Banking application, how would It get a query result set like Customer Account Transactions for a month from Queue, the request being fulfilled by an ERP System? Does the web application need to have a service listening to a queue all the time?
2. How does the service call by the Application (web based, desktop etc) need to be structured? For example, in the case of web based application if it calls a service to return the balance of Customer account in one of its Method/Function, does the function need to make a synchronous call to the service and wait for the result (customer balance) to end the function call execution. In other words, how does the web application need to poll on the reply queue.
Regards,
Sajid. |
|
| Back to top |
|
|
| fjb_saper |
| Posted: Fri Oct 06, 2006 2:50 am Post subject: |
|
|
Grand High Poobah
Joined: 18 Nov 2003
Posts: 20756
Location: LI,NY
|
Read the manuals, search the WEB for JMS patterns. You will learn more and way faster than by reading some answers you may not understand here.
We are talking about industry standards here...
Enjoy
_________________
MQ & Broker admin |
|
| Back to top |
|
|
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
|
|
