Correlation with request response port

I created a correlation type and correlation set on the behalf of File.ReceivedFilName context property.
I am getting the initializing correlation set property in the send port as follows:
But in Receive Port I am not getting the Following correlation sets property as follows:
Why this is occuring ?
Am I doing something wrong ?
Is correlation supports with request response port ?
Prakash

Hi Prakash,
When you drop the Receive shape in your Orchestration, you would see both the “Initializing Correlation Set” and “Following Correlation Set” properties for the Receive Shape (same case for the Send Shape).
When you have linked the Receive Shape to the Request-Response port in your Orchestration (Request-Response port whose Port direction of communication is set to “I’ll be sending a request and receiving a response” property), the “Following
Correlation Set” property for the Receive Shape disappears (same case for the send shape). Because you have connected the send/Receive shape to the two send port and BizTalk can in turn handle the correlation. You don’t need to specify the correlation set
properties explicitly. These properties need to be configured only when you configure your send/Receive shape to one-way ports where you need to instruct BizTalk on how to handle the correlation not when you configure send/Receive shape to two way send port.
If this answers your question please mark it accordingly. If this post is helpful, please vote as helpful by clicking the upward arrow mark next to my reply.

Similar Messages

Expose an orchestration as WCF service with a Request-Response port set to Direct Binding

Hello ,
I'm trying to expose an orchestration as a WCF service (while the Request-Response port is set to Direct binding) . Is this possible ? (the issue will be correlating back the response to the synchronous WCF request after the response is posted back
to the Message Box by the Direct Send logical port).
I have no problem in exposing the orchestration while I'm setting the Binding port to 'Specify Later' (What I'm trying to do actually is to expose a orchestration using different adapters (like Http and Wcf ) in the same time ).
Thanks

Yes, you can publish an orchestration as WCF/Web service which has direct bound request-response port.
Subscription works based on “EpmRRCorrelationToken”
and “RouteDirectToTP”. When you enlist the orchestration which
has the direct bound request-response port (also published as service), its subscription are based on above two context properties and they take care of correction.
And also the response matches to the request part of the port due to the nature of two-way communication protocol.
Create an Orchestration with direct-bound request-response port and publish it as service. This will work. Give it a try
Regards,
M.R.Ashwin Prabhu
If this answers your question please mark it accordingly. If this post is helpful, please vote as helpful by clicking the upward arrow mark next to my reply.

BAM - using with Request Response Send Port

Hi,
I have a scenario where a message is received on a receive port, sent to a solicit-response send port, and then the response is sent out on another send port. All ports have failed message routing enabled, so that any errors are sent out to an exceptions
send port.
I am interested to hear options on how you would set up a BAM tracking profile for this in the TPE to log the time received, time sent on the two way send port, time response returned and then finally the time the last send was carried out? Would you
use a single Activity/View and continuations, Multiple Activities or Views and multiple tracking profiles etc? How would you include the time that an error was sent to the exception port.....
I have a solution, but I am not sure it is ideal and I would be interested in how others would tackle this. If you know BAM, feel free to offer an opinion :)
Thanks
Stu.
Stuart Brierley - BizTalk Integration Specialist - http://geekswithblogs.net/StuartBrierley/Default.aspx

Hi Stuart,
You can start up your activity from Orchestration .No need to have a continuation from Pipeline as you can use xml receive .
You solution is simple were you are receiving your message from location and hitting a solicit-response
port .
Just have a Expression shape to start up with your activity and on the points when you want to update
the activity.
Code is listed below
//Create a Activity ID
string activityId = Guid.NewGuid().ToString() ;
//Start the activity with your Activity Name
Microsoft.BizTalk.Bam.EventObservation.OrchestrationEventStream.BeginActivity(“SampleActivity”, activityID);
// Updates the activity record.
Microsoft.BizTalk.Bam.EventObservation.OrchestrationEventStream.UpdateActivity(“SampleActivity”, activityID, “ReceiveLocation”, Rcvmsg(BTS.Receivelocation), "TRANSACTION_CREATED", DateTime.UtcNow);
// End the activity record.
Microsoft.BizTalk.Bam.EventObservation.OrchestrationEventStream.EndActivity(“SampleActivity”, activityID);
If you want to work with same activity in continuation with Other Orchestration (Generic Exception Handler may be ),then create a continuation token with a Property schema and from parent Orchestration assign value to it.
continuationToken = "_Continuation" + msg_Internal.ContinuationID;
Hope this will clear your ideas as you dont require to start up tracking from Pipeline .
Thanks
Abhishek

OSB with request response MQ business service

Hi,
We have OSB webservice whice interacts with MQ via business service. We need to send the request to MQ request queue and get the response from the response queue.
For this, we have configured a business with MQ transport, request-response is enabled, message id is set for the correlation, auto generate correlation value is chosen.
The business service sends the message but unable receive the response from the response queue. Failing with the error message like
The invocation resulted in an error: [WliSbTransports:381918]Failed to receive response, within the configured timeout, for request message with id 42454BD3E5F7647914544E793F9F0000013EC9DF03918057 and correlation id 42454BD3E5F7647914544E793F9F0000013EC9DF03918057.
But we checked with the end system, the response is sent tot he response queue in no time. We also increased the time out but still same issue.
We noticed that if we dont send the message id, we are getting the response but not response of the request sent but some other response.
Please suggest what needs to be done to fix the issue.

Open your business service and navigate to HTTP Transport configuration page and check what is the http type is enabled.
Look to me its with GET method, change it to POST and re-try.
If the above solution is not helping, try to check the Follow HTTP redirects check box below the Advanced Setting in same page.
Thanks,
Vijay

CDATA section ignored at request-response port

Hi Friends,
I am calling a request-response service from BizTalk WCF-Basic Http port.
The request consists of CDATA section and service behavior is defined to return a response with similar structure.
<Request>
<![CDATA[<RequestID>123</RequestID>]]>
</Request>
The expected response is -
<Response>
<![CDATA[<ResponseID>789</ResponseID> <ServiceID>963</ServiceID>]]>
</Response>
I tested the service with SOAP UI and it gives desired results as above.
But, when same service is invoked through BizTalk WCF-Basic Http port, response received is -
<Response>
<ResponseID>789<ResponseID><ServiceID>963<ServiceID>
</Response>
But why BizTalk is behaving like this? Looks like BizTalk is ignoring CDATA section and returning string.
Is it possible to apply some configuration at port something similar to
disable-output-escaping=”yes” , so that port does not convert CDATA to string?
Hope someone can help!

So where is the problem in this behavior? The reason the service is expecting the request and response as CDATA elements is because it saves them from having to change the endpoint schema every time they release a new request response.
When you create the schemas generated by consuming the service promote the "Request" and "Response" elements. Then in the construct shape you can populate these using the promoted elements and assign the xml.OuterXml.
Similarly while processing the response you can load it into the xml using the XmlDocument.LoadXml(serviceMessage.Response);
Regards.

How to work with request, response and sessions in JSC

Hi
I have two Servlets ServletA and ServletB. ServletA creates a session and add some attributes to this session. Then it forwards the request to ServletB where the attributes of this session are retrieved and used in response object.
Here is what I am doing in ServletB
- Get the session attrribute. This attribute is of type byte[]
byte[] mydata = (byte[]) request.getSession.getAttribute(data);
response.setContentLength(mydata.length);
OutputStream stream = response.getOutputStream();
stream.write(mydata,0,mydata.length);
stream.close();
My question is ...
How do I achieve this in Creator. ?
Thanks in advance
Srinivas

Hi
I have two Servlets ServletA and ServletB. ServletA
creates a session and add some attributes to this
s session. Then it forwards the request to ServletB
where the attributes of this session are retrieved
and used in response object.
Here is what I am doing in ServletB
- Get the session attrribute. This attribute is of
type byte[]
byte[] mydata = (byte[])
request.getSession.getAttribute(data);
response.setContentLength(mydata.length);
OutputStream stream = response.getOutputStream();
stream.write(mydata,0,mydata.length);
stream.close();
My question is ...
How do I achieve this in Creator. ?
Thanks in advance
SrinivasBy "this" do you mean "how do I access a session attribute that has been placed there by another servlet?"
The base class for all page beans has a bunch of convenience methods available. The simplest one to use for this purpose is getBean(), which searches through request scope, session scope, and application scope for a bean with the name you specify -- and, if its a managed bean, will create it for you as well. So, your code would look like this:
String data = ...; // Name of the attribute you want
byte[] mydata = (byte[]) getBean(data);There is an additional quirk related to what you're actually trying to do, though ... it looks like you are trying to write out binary content (perhaps an image loaded from the database or something). It is not possible to mix binary and text output in a single response.
Your best bet for this particular purpose, then, is to continue using ServletA and ServletB just as you have been. The applications created by Creator are standard servlet-based web applications, so you can use additional servlets and all the other goodies (although you'll have to declare them yourself in the web.xml file, as usual).
Craig

EDI over WCF request-response port - how to catch failed messages and send response over a 2 way port

I'm using BizTalk 2009. I've written an orchestration for accepting 270's and returning 271's, and another for 276's and 277's. The incoming and outgoing EDI stream is wrapped in XML. I've published the schema using the WCF publishing wizard,
and I am using the WCF-WSHttp adapter. I have custom pipeline components to unwrap and wrap the XML, and to handle the NACK generated by the adapter. Everything works fine if valid EDI is received.
However, I'm challenged with handling failures from the EDI disassambler. The specification I've been given is to return a 271 or 277 in all cases (or as many as possible). What I would like to do is catch the failed message, generate a (271
or 277) response, and return that response over the 2-way port. I know how to enable failed message routing and how to subscribe to the message if it were to come in over a one-way port. It seems that this cannot be done when the message comes
in over a 2-way port.
Any suggestions on how this can be done?

I should have mentioned that turning off as much EDI validation as I can in the pipeline is already part of my approach. However, to my knowledge, there are some validations (such as incorrect segment count in the SE) that cannot be disabled.
Regarding the TA1/997/999 - because this is a real time transaction, I'm not generating any. Still, to take your suggest, I decided to generate a 997, thinking I could at least subscribe to it and if the 997 indicated a failure, to
generate the response (in the form of a 271 or 277) that my trading partner is expecting. What I found was that the 997 was being returned through the send pipeline of the 2 way port, EVEN WHEN I set RouteAckOn2WayPort to False on the receive pipeline!
This produces even more work for me, because now I have to "eat" the acknowledgement so that the 271 or 277 can get through.

Request Response (database) biztalk Error hadling

best was to handle exceptions working with Request Response port ? hitting database.
1) what properties should be set in Scope shape ?
2) what should be set the Transaction type of Orchestration ? when to use Long Running or None ?
3)does on ReqResp SP, transmitted set to true works ?
MBH

Well, you've already answered your own question, use an Exception Block.
What exactly do you mean by 'no response'? There are two realistic possibilities, A) a timeout occurs or B) no result set is returned.
(A) is an error condition and the exception will be raised back to the Orchestration and caught in an
Exception Block.
(B) is a real possibility depending on how the SP is coded. The operation technically completes correctly, but no result is considered by the caller to be an error condition. This you would have to test for.
If the database code silently fails, as you describe, that is not your problem. The owner of the Oracle app has to fix that since you can't know if an error occurred or not.
You have to try these things. It's not possible to offer any more meaningful advice because exactly what you have to do is completely dependent on how your Oracle instance behaves and what the business case considers an 'error'.
Please, set it up and try different scenarios against Oracle and tell us what happens. If something doesn't work or doesn't meet expectations, please open another thread for that specific issue.

Messaging only: WCF-BasicHTTP adapter Request-Response correlation

Hi,
we have a solution where we make a soap web service call to a third party.
we do not want to use orchestrations.
we are in need to pass a value to the third party service and we would
like to have that value passed back to us in the response.
However, the third party is not guaranteeing this.
as an alternative, i tried promoting a property into the context
before the outbound call and check if this is still available in the context when the response come thru.. but no luck here.
IS there any other way to achieve this?
regards,
MS

Hi MS,
What is your question? When you use Request-Response (Receive) port and solicit-Response (send) port, the correlation would happen out-ob-box with the help of EpmRRCorrelationToken, CorrelationToken and RouteDirectToTP
properties.
Are your not using the solicit-Response (send) which would take care of the correlation to Request-Response (Receive) port. Ensure you're using the XML-Transit (on send) and XML-Receive (on receive) pipelines on these two way ports (in both Request-Response
(Receive) port and solicit-Response (send) port)
If this answers your question please mark it accordingly. If this post is helpful, please vote as helpful by clicking the upward arrow mark next to my reply.

Incomplete request/response operation missing send

Brief on orchestration whihc have request response port(published as WCF service)
1. receive(initialize co-relation)
2. Looping and do some operation(in scope and exception handling)
3.listen with receive(follow co-relation )& dynamic set delay shape(this step is also in loop, delay continues untill second request is received)
4. construct resp message and send
getting build error incomplete request/response operation missing send

You have a loop shape, and obviously an condition associated to that loop.
Now consider a scenario: you receive a message - and your loop condition is not met. In such cases you are coming out of loop (after receiving a message), and you will not be able to send anything back on the Req/Res port. Hence the Compiler restricts you
this way.
Thanks
Abhishek

Request Response : Direct bindings

Hi all,
I want to expose a schema as a webservice in BizTalk, but my orchestration instead of having a
specify later port bounded to my receive location, it's bounded
directly to the message box how can i handle that by consuming the message request
and sending to the same Receive location my response message ?
It is possible in BizTalk to use a request response port in direct binding mode ?

Thank's for your reply. But you didn't explain how to route the response message to the same receive location. If you try you will get an error because no subscribers are identified to consume the response message.
There must be a mechanism to implement, but i do not know how ?
And without adding a filter just configure the request part to the message type exposed and it is done.

BlazeDS RPC request-response - Require ability to push server-response asynchronously

Hi BlazeDS developers,
I understand that BlazeDS has RPC req-response async client and server mechanism (whats the big deal html UI also has been aync?). It also has publish subcribe pattern with ability to have server side push to client.
My questions is if below ability present in BlazeDS in request-response RPC?? (Do not want to use pub-sub messaging.):
Asynchronous mechanism for a response "at server side" (not between client and server) i.e.
1. a request from flex client using RemoteObject on a POJO at server,
2. POJO at server processes the request asynchrnously, i.e. spawns a thread, and does some business logic calculation.
3. Has the response ready after few seconds.
4. prepares the response and pushes to client
Server business logic does not have to block client POJO object function call made from RemoteObject RPC from client. (consider RemoteObject use case)
Thanks in advance for your comments/answers
Nilesh

Hi Nilesh,
What you describe is not currently possible in BlazeDS with request/response RPC requests. While handling of remote object requests is async in the client code, it's not async between the client and the server.
A bug has been logged to get this functionality added and it is currently targeted for inclusion in the next release of BlazeDS.
Here is the link to the bug if you want to watch its progress.
https://bugs.adobe.com/jira/browse/BLZ-184
Hope that helps.
-Alex

Request Response Receive Port and Solicit Response Send Port

Can anyone give me some idea how to bind Request Response Receive Port and Solicit Response Send Port with each other through TCP/IP adapter ??

The binding of a Receive and Send Port has nothing to with the choice of Adapter.
1) There is no out-of-the-box TCP/IP Adapter for BizTalk, but several 3rd. Party Adapters has been made, along with this one in CodePlex: http://tcpipbz2010.codeplex.com/
2) Most people will use an Orchestration for binding a SR Send Port to a RR Receive Port. This may not be necessary though. All you need to do is have your Send Port subscribe to your Receive Port (by setting the Filter on the Send Port to "BTS.ReceivePortName
== [The name of your Receive Port]". In this case, internal correlation will make sure that the response is routed back to the Receive Port.
Morten la Cour

Publish/subscribe in a request/response manner with WCF?

Is it possible to make WCF service work in request/response manner using WCF duplex channels?
Or is there any kind of automatic correlation between the messages when using WCF?
I want to be able to attach WCF service to a topic, and when a request comes, the service to return result, that will be transformed to brokered message and returned over the bus to the caller.
Thanks in advance!

It could help, If there is a way to bind a relay endpoint to a topic...
The scenario in details: I have a service that extracts an object from a database by given identifier. I want to be able to activate that service with a message over a topic. That part is OK, I've did it. But I want the service to return the extracted
value over the topic (or another topic) but I do not want to change the service contract to be OneWay - for interoperability, I need the method in question to take one argument and return object.
So that's why I am searching for a way to bind one message to call the service's method and when it is ready to transform the result in another message and publish it back on the bus.

Received response from host (router IP address) with invalid source port 32784

I replaced my old wireless router with a Cisco Linksys E4200, running firmware version 1.0.02 build 13 May 24, 2011. About once a minute the router sends an unsolicited DNS message to the IPV4 multicast address 01:00:5e:00:00:fb with a destination IP address of 224.0.0.251. The unsolicited message is a DNS response with source port 32784, transaction ID 0, flags 0x8400 (standard query response, no error), questions 0, answer RRs 2, authority RRs 0 and additional RRs 1. The two answers both relate to the router itself: one has Name Cisco18738.local, type A (host address), class 1 (IN), cache flush true, time to live 1 minute, data length 4, and the address of the router. The other is the reverse of the same address. The additional record is for Cisco18738.local, type NSEC, class IN, cache flush true, time to live 1 minute, data length 5, next domain name Cisco18738.local, RR type A (host address).
When my desktop computer receives these messages it logs an error, for example: "Jun 23 07:39:22 sauterws02 avahi-daemon[1067]: Received response from host 10.146.9.1 with invalid source port 32784 on interface 'eth0.0'" The 10.146.9.1 is the router's IP address. I also see these messages on the wireless link from my laptop.
I suppose the E4200 is generating these DNS messages in a misguided attempt to make sure there is no old information about its name. Is there a way to turn them off? If not, is there a way to report this to Cisco as a bug?
Solved!
Go to Solution.

gv wrote:
1. To contact Linksys, call support.
2. From the internet draft: "Multicast DNS implementations MUST
silently ignore any Multicast DNS Responses they receive where the
source UDP port is not 5353." Your avahi-daemon does not comply with this draft.
Thank you for the reference. For the sake of others who may read this thread, the current draft of multicast DNS is at http://www.ietf.org/id/draft-cheshire-dnsext-multicastdns-14.txt.
Here is the whole paragraph from which you quoted, from section 6 (Responding): "The source UDP port in all Multicast DNS Responses MUST be 5353 (the well-known port assigned to mDNS). Multicast DNS implementations MUST silently ignore any Multicast DNS Responses they receive where the source UDP port is not 5353."
Thus, it appears that there are two errors here: the Cisco E4200 is not using 5353 as the source port, and the logger is not silently ignoring it.
The message appears to be an announcement rather than an actual response to a query. In section 8.3 (Announcing) I found this paragraph: "A Multicast DNS Responder MUST NOT send announcements in the absence of information that its network connectivity may have changed in some relevant way. In particular, a Multicast DNS Responder MUST NOT send regular periodic announcements as a matter of course." Thus, it appears that there are three errors here.

Correlation with request response port

Similar Messages

Maybe you are looking for