BPM fork problem

Hi,
I am working in a scenario where I need to receive messages from multiple publishers on different fork branch and sequence it. I have configured the scenario as mentioned in basis namespace example. I but it looks like even in parallel branches fork expects one of the receives step needs to be configured for start process and remaining with start process flag unchecked. If I mark all the receive process as start process it expects only the last branch to be sent first. If other message arrives first it shows scheduled to be delivered and doesnu2019t get processed.
Did any face this issue? Any help appreciated in advace.

Hi
This should not happen. If you mark all receive steps as start of process then if any message will arrive it will trigger the BPM and then other will receive to same BPM instance based on correlation.
Check correlation you defined for those receive steps are correct and including all three messages in correaltion.
One other reason also can be queue to show ur message in sheduled state. Refer folllowing pdf:
[https://www.sdn.sap.com/irj/sdn/go/portal/prtroot/docs/library/uuid/e0ff0006-3901-2a10-9e97-a71b423da545]
Regards
Sami
Reward points if helpful.

Similar Messages

BPM Fork Processing without receiving all required messages

Hi
I am developing a BPM which receives 5 different messages.
I am putting a block with Local Correlation with Block Start = New Transaction and Block End = New Transaction.
Inside this block, I am using Fork with Necessary Branches = 5 , using local correlation for each receive step with Use Correlation and Activate Correlation.
When I am adding Receive Steps to Fork, there is an extra branch getting added to the Fork. So now I have 5 Receive steps with Abstract Interfaces and one extra step without having anything. I couldn't figure out how to get rid of this extra step. I know there is way to get rid of this extra step in Fork...but not sure how
Now, when I am testing, after receiving any 4 messages process is coming out of the Fork block and going to the next block.
As in the next block I use all the 5 messages, I am getting error there.
Can you please give your inputs what could cause this problem?
How do we get rid of extra branch in Fork which we do not use..
Thanks for your help in advance.
Regads,
Ramesh

Hello Ramesh,
When you Drag-Drop any process step in the Fork, it will create a new branch and will sit inside that branch. If you want to put the step on some specific branch then Drag the process step and take it to the desired branch and only drop when [ ] appears on that branch. To insert new branch in fork you can use Right click on Fork->Insert->New Branch.
Now for ur problem, you have set the necessary branches as 5, when u receive the 4 message(4 branches executed) and fifth branch(empty branch) will make total count to 5. Therefore, ur fork step will be executed and will not wait for the fifth msg.
HTH,
Regards,
Sami.

BPM giving problem in Merging and Splitting

Hi Gurus,
Im trying to BPM scenarios,
First i tried to Merging Two message and send to one receiver system. In that both Sender and Receiver communication channel are wokring fine.. all are showing green light.. and the files also got processed from sender side but i cant get result the output side. i checked my abstract interfaces , interfacing mapping and Container variables in IP everthing is right.. Even this is not 1:N or N:1 mapping , all message types are having occurance of 1:1 only. when i checked using SXMB_moni... there my two sender are succesfully sending the messages and ip getting the message.. after IP should act as a sender right. that msg not there... what would be the problem... even correlation also correct using the common filed's xpath i was combing the msg. In IP im getting the Yellow color warnin like correlating not used in sender step but no error. but i activated the correlation in sender step and assigned the xpath of the common field.
After that i tried to do splitting same problem occuring.. no message in receiver side.. i double checked my abstract interfaces, inbound , outbound interfaces, interface mapping and container variable everthing was assigned properly. In IP Only a blue color waring i getting like receiver determination initialized but not used but no error , but when i checked the interface determination it was fine.. i created receiver type local variabled with multiline . this is not 1;N or N;1 mapping... a simple mapping
Please help me experts,
Regards,
Balaji

Hiii experts ,
Please help me.. i have tried many times.. still the same problem exits...
source msg type having 5 elemtn.. first 2 element mapped first target msg type and remaining fields mapped to receiver 2 target msg type.
IP step are configured as below
Container Variables: send,recv1, recv2, 2 receiver type msg for receiver determination (Receiver1, Receiver2) both are multilined.
Receiver step : sender msg type (sender abs interface type)
Transformation : Has the interface mapping and one source msg and 2 target msg
In fork
Branch1:
receiver determination1 :receiver msgtype (recv1) and the multiline element (receiver1) type of receiver.
sender1 step: receiver msg type (recv1) and Receiver from-> Receiver List : Receiver->Receiver1
Branch2:
receiver determination2 :receiver msgtype (recv2) and the multiline element (receiver2) type of receiver.
sender2: step.. receiver msg type (recv2) and Receiver from-> Receiver List : Receiver->Receiver2.
But msg going to IP.. after that IP not start to send msg...
Please tell me.. what im doing wrong here?
Regards,
Balaji

BPM BpmPatternCollectTime problem

Hi,
I've been trying to get the example pattern BpmPatternCollectTime from SAP to work on PI 7.1. This pattern waits for messages and collects them into one message. I've checked my BPM against the SAP BPM and the only difference is that I'm using XSLT mapping in the Transformation step. Is this causing the problem ?
I've checked all the blogs I could find, searched for SDN messages, checked SXI_CACHE etc. to no avail.
Scenario is: I'm sending multiple IDOCS from SAP HR system via PFAL, each has 200 persons in them.
I map the IDOCs to the target format (XML) and collect them in the BPM and send to file adapter.
Problem is I only get the first IDOC contents in target message, and the rest are missing.
When I look at the BPM workflow log, Container tab, I can see that there are 3 messages in the MESSAGE_IN entry. I can open them and see the contents fine. This leads me to believe that my Correlation is working ok.
Also in the Block step the COLLECTMESSAGELIST contains 3 entries as it should.
Correlation says:
<CorrelationKey Version="001"><ID>SAP HR</ID></CorrelationKey>
Where 'SAP HR' is the content of the correlation XPath query I defined.
MESSAGES_OUT in the BPM workflow log contain the same 3 messages.
TRACE in the BPM workflow log shows that the mapping of the Transformation step executed ok. Transformation step Operations Mapping is using an XSLT mapping that is copying the input to output as is.
When the BPM is finished the target message only contains the first IDOC data.
How can I get the rest of the data to append to the target message ?
Kind Regards,
Jari Pakarinen

Hi,
Using this mapping XSLT I was able to get the BPM working:
<?xml version="1.0"?>
<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" xmlns:ns="http://xxxx.com/xxx" xmlns:ns0="http://sap.com/xi/XI/SplitAndMerge">
<xsl:template match="/">
    <ns0:Messages xmlns:ns0="http://sap.com/xi/XI/SplitAndMerge">
      <ns0:Message1>
        <xsl:copy-of select="*"/>
      </ns0:Message1>
    </ns0:Messages>
</xsl:template>
</xsl:stylesheet>
This means you have to wrap the mapping inside
<ns0:Messages><ns0:Message></ns0:Message1></ns0:Messages>
tags. This took a while to find out.
Jari

BPM 11g: problem initiate task from API

Hi all,
First of all, i have deployed a BPM process and his initiator task is a human task implementation. Then the user can click on it in teh workspace and they are redirect to my J2EE aplication, in wich i want throw a new instance of my process when the user complete the form data.
I'm do know how to implement it, and i trying two ways:
one: initiate a process instance with oracle.bpm.service. package the the server give me this error:
http://kr.forums.oracle.com/forums/thread.jspa?messageID=9244238
java.io.InvalidClassException: oracle.bpel.services.workflow.verification.impl.WorkflowContext; local class incompatible: stream classdesc serialVersionUID = 3642712178109525089, local class serialVersionUID = 1
two: initiate a new task programmatically, is it wright? a new task of my human initiator task from java is that i need for create a BPM Process? i don't know, becouse i get this other error:
http://download.oracle.com/docs/cd/E15523_01/integration.1111/e10224/bp_worklistcust.htm#CIHEAFBB
WSM-09004 Component auditing cannot be initialized
any ideas? thank you very much in advance, and sorry for my poor english.

I hope you have got the solution for this problem.
will you please share the solution. as I am also facing the same...
Thanks

BPM Worklist Problem After Setting Frontend Host and Port

Hi all,
I was following the EDG for SOA 11.1.1.5 and was encountering a problem with the BPM Worklist after setting the Frontend Host and Port. We have a topology that includes a load balancer that terminates SSL, two OHS instances, and two SOA instances on separate VMs. When BPM Worklist was loading, it was calling webservices internally through the load balancer, but WebLogic was expecting them as http://. It was not able to resolve the HTTPS port that was being sent from the load balancer through the 80 to 443 redirect.
What the current EDG does not tell you, is that you have to also turn on "Enable Weblogic Plug-In" for WebLogic to use the OHS plug-in. I did it at the domain level, as all traffic will go through the load balancer. This essentially tells WebLogic that all URLs are https://
Here is an exerpt from the SOA.out log file:
<Sep 8, 2011 9:32:06 PM PDT> <Error> <oracle.soa.services.workflow.worklist> <BEA-000000> <<.> Service error.
Internal Error; Service error occurs in IdentityService in method lookupUser.
Refer to the log file that is configured for oracle.soa.services.identity for more details on this error and contact Oracle Support Services
ORABPEL-10585
Service error.
Internal Error; Service error occurs in IdentityService in method lookupUser.
Refer to the log file that is configured for oracle.soa.services.identity for more details on this error and contact Oracle Support Services
Caused By: javax.xml.soap.SOAPException: oracle.j2ee.ws.saaj.ContentTypeException: Not a valid SOAP Content-Type: text/html; charset=UTF-8
What it is also missing, is that you should also set the SOA Infrastructure ServerURL mbean for the load balancer, to match the frontend host and port. Do this through Fusion Middleware Contol. Otherwise, you could encounter URL mismatches.
I hope this helps someone else.
I've asked Oracle to add this to the EDG for SOA. It's in the IDM guide, but not in any other EDGs.

Thanks Josh. It helped me infact in 11.1.1.4 Enable Weblogic Plug-In is not required. But 11.1.1.5 I thing it is mandatory. Oracle should have a clear documentaion.

Sync/Async BPM scenario : Problem at Integration Process..!!

Hi All,
I am doing Sync/Async BPM scenario.
start ->Receive(Request)> Transformation(Responsemap)>Send(SendtoR/3)->Send(Send Response) ---> stop.
When Iam trying to Activate, Integration process giving errors as
Source parameter AA_Payment_Req | http://XXXXXXXXX.com/SAP_PI/DirectSales/Payment not set Target parameter AA_Payment_Response | http://XXXXXXX.com/SAP_PI/DirectSales/Payment not set
Warning: Container element Response is defined but is neither initialized nor used.
But I am using them in the BPM. The same kind of scenario is working for another interface which is already running in production. When I tried to check that interface it is showing only Warning message as Container element Response is defined but is neither initialized nor used.
I tried deleting and creating again and still it is showing the same error.
Can anyone please suggest what might be the problem.
Regards
Deepthi

Hi,
could you pls let me know how its solved, while i am trying the same i am also getting the same problem..
could you pls explain bit more on this.
Thanks,
Venkat.

BPM Cache Problem

Hi All,
I hav created a Business process. when i modified that process, it is not getting reflected in the SXI_CACHE. No matter how many times i activated it, it is not getting modified. i tried for around two hours and restarted the server again as the last alternative. I had refreshed the cache even with the mode 'full' which din't result in the expected outcome.
Any suggestions about this as i don't want spend two more hours next time.
Thnx
Anil.

Hi guys,
I had done the samething which u ppl were mentioning for two hours. but my problem is that the BPM is taking the previous version itself.And i din't know how to enable BPM to consider the latest version.And now, my problem got solved.
In the Integration directory, i deleted my integration process(say MyXXXBPM) and created a new integration process with the same name(MyXXXBPM) and imported the same process. It is taking the latest version now.
Anyways.......Thnx guys
Anil.

ADF-BPM integration problem.

Hi,
1.I have a small scenario like integrate oracle ADF screen with Oracle BPM Human Task. we didn't use auto generate task form(we are using one adf application which is already developed in Oracle Adf technology).
2.Those adf screen is populated in BPm work space using url of ADF screen placed in em console.It's fine,it will populated in work space.
3.But my problem is we can click on the buttons approve or reject on populated adf screen then only human task will modify the running state to completed state.
Any one have the idea on my requirement Please give the reply.
(OR)
Is there any other option can u please post the solution.
1. My actual scenario is integrate oracle adf Screen with Oracle BPM humantask.(already devloped. note not an auto generate adf application).
2.That screen Would be populated in BPM work space.So user login to BPM Work space And click on the tabs Reserved or cancel on populated adf screen.
3.Then only the human will be the state like running to completed state.
Edited by: Alluri.S.Raju on Dec 6, 2012 12:45 AM

Hello Rudra Raju,
Did u get the solution...
Even iam looking for the same..
Also I need to pass the Employee ID from Human task to ADF UI So that the User can download the file based on the Employee ID.
ThankQ!!

BPM "Fork" step question

Concerning fork step : Here is the documentation :
You use a fork () when you want to continue a process in branches that are independent of each other, for example, to communicate with two systems that are independent of each other. The branches of the fork join in a union operator.
You can specify the required number of branches and then define whether the process must run through all branches, or just a particular number of branches. Furthermore, you can define an end condition for the fork (see also: Condition Editor).
As soon as a branch reaches the union operator at runtime, the system checks the following conditions in the specified order:
·        The process has run through the required number of branches
·        The specified end condition has returned true
The step is complete as soon as one of the conditions returns true.
I want to use it to do 3 different send process in parrallele with 3 different receivers.
Before Fork step I use a container to initialize varaiable count = 0
In each branch I do my send step and then a container that does count = count + 1
And finally I have defined my fork with following end conditions :
End condition : count = 3
Necessary branch : 3
But my BPM allways executes only one branch !!!
(Only one is in green in the graphical editor of monitoring for BPM)
Any idea ?
Do I use the right solution to do the following :
"You can specify the required number of branches and then define whether the process must run through all branches, or just a particular number of branches"

Hi Jean-Charles,
I agree with Michal. Is it really necessary to use a fork for different send steps.
The reason for your unexpected termination could be the condition. Each fork branch reaching the union operator checks here in this sequence:
Process has run through required branches (here: 3)
End condition is true (container = 3, you have the same container element for ALL branches? This should not work. You need one for each, I think.)
Step is complete as soon as ONE of the conditions return true.
So use three different container elements as counter and connect them in the end condition with AND.
Regards
Dirk
Message was edited by:
        Dirk Meinhard

Urgent : XI BPm Performance Problem

Hy All,
we have following big performance problem with BPM Scenario.
Of the inbound site we can get at the same time more than 10000 messages that have to be read from BPM queue. The BPm execution time in strongly affected and slowly. Could anyone tell me how to improve the performance (parallel queue execution.....).
I hope in your help.
Thank's a lot
Mat

Hi Mattia,
see for improving performace, you need to restructure your BPM process....if you want your input messgae to be processed in different queues, this is not possible if all your message are of same message type as then XI uses the same queue for all your messages..........so see in your scenario if BPM is really needed or you can do your scenario without BPM.......
Thanks,
Rajeev Gupta
Message was edited by:
RAJEEV GUPTA

New BPm Definition Problem

Hy all,
ich have the following problem. I made modification in my BPM. The modifications have been detected but in a multimapping Step (old Version 9:1 messages -> new version 10:1 messages) it continue to use the old Version of the mapping. I try to create one more time the mapping step but with any result.
Could anyone help me.
Thanks a lot
Mat

Mat,
Can you try this.....
Delete the interface mapping, Since you have created a new mapping program now activate the change list. Again create the interface mapping. Activate the changelist. In the BPM delete the trnasformation step which has the multimapping part. Activate and again insert the transformation step. Now try to use different container variable. Activate the changes list.
The point is at every step of creation/change/deletion activate the change list and check if it works. Also avoid using the same naming conventions for the objects.
thkx
Prabhu

BPM Queue problem

Hi there, i have a BPM with a Block step that has an infinite loop and collects delivery idocs, i have an exception set to 2 minutes or 500 idocs, basically when i send more than 500 idocs or it takes more than 2 minutes the first group is processed perfectly, but the remaining idocs are stuck on a queue visible under the smq2 with the status READY, does anybody has an idea of what could be the problem?,
Best Regards,
Roberto.

Hi there,
i cannot see any errors, i do have a correlation active by some region ID, but my test involves only delveries of 1 region, i don't know what could be wrong about the correlation, in addition the bpm steps goes as follow:
- infinite loop with a receiver step inside a block step with a timeout exception
- after that i transform those collected messages into 1
- synchronous call to a 3rd party via webservice
- send back the response to ECC
Roberto.

BPM Correlation problem

Hi All,
I did BPM correlation scenario. According to this scenario I have to send 2 messages, if I send two messages from the sender side it is working fine. If I send one message from sender side it is waiting until second message reach. Here my problem is if the second message may not send, I want to kill the message waiting process at a specified time. Please help me how to do this process.
Thanks ,
Kumar.

Hi,
There is a 'Deadline branch' dedicated for this purpose. In a block you can introduce a deadline branch.
Go thro the link for more details:
http://help.sap.com/saphelp_nw04/helpdata/en/c5/582541019fa52fe10000000a155106/content.htm
Regards,
P.Venkat

Windows Sharing and resource fork problems

Hello, I am into my second month of Mac ownership and today I tried to enable Windows Sharing to allow a PC here to map to the Mac and share files.
I enabled the Windows Sharing feature in the System Preferences. Then I tried to map to the mac using the PC (Win XP). The PC sat there for a while (maybe about 30 secs or so) trying to connect, then displayed an error message saying a semaphore timed out (I don't remember it exactly).
Anyway, I went back into System Preferences and all the options for Sharing were greyed out. I couldn't turn off or turn on anything!!
So I looked in the console log and found the following errors:
2006-08-30 18:12:40.119 System Preferences[15066] [NOTE: this exception originated in the server.]
couldn't fork
/sbin/service: fork: Resource temporarily unavailable
2006-08-30 18:12:43.336 System Preferences[15066] [NOTE: this exception originated in the server.]
couldn't fork
/sbin/service: fork: Resource temporarily unavailable
2006-08-30 18:12:52.009 System Preferences[20541] hardwareCheckThread got an exception name:NSInternalInconsistencyException reason:couldn't fork
Aug 30 18:13:37 localhost launchd: org.samba.smbd: fork() failed, will try again in one second: Resource temporarily unavailable
2006-08-30 18:16:53.360 System Preferences[21610] hardwareCheckThread got an exception name:NSInternalInconsistencyException reason:couldn't fork
2006-08-30 18:19:13.214 System Preferences[21611] hardwareCheckThread got an exception name:NSInternalInconsistencyException reason:couldn't fork
2006-08-30 18:31:48.547 System Preferences[21636] hardwareCheckThread got an exception name:NSInternalInconsistencyException reason:couldn't fork
2006-08-30 18:54:23.310 System Preferences[21650] hardwareCheckThread got an exception name:NSInternalInconsistencyException reason:couldn't fork
In the System log there are a whole bunch of these messages:
Aug 30 18:11:23 localhost kernel[0]: proc: table is full
I went into the Terminal and tried to get to a command prompt but it just sat there with a blank screen.
So I shut a bunch of apps I had open down only keeping Firefox open and I searched online for issues regarding this error - found information telling me I could kill the smbd processes and up the max proc limit.
After waiting a while (didn't want to reboot - I'm stubborn :-)), reopened Terminal and could see the command prompt. But anything I tried to do regarding killing the smbd processes or even doing a directory listing or a process listing gave the error: fork: resource temporarily unavailable
So I just waited it out - eventually I was able to kill the smbd processes. After I did that, I went back into System Preferences and was able to work with the Sharing preferences again (I disabled Windows Sharing).
And after all that, here is my question(s):
- is this problem going to keep happening when I turn on Windows Sharing and a PC tries to map to the Mac?
- do I need to up the process limit if I'm going to turn on Windows Sharing?
- is there a bug/issue I don't know about regarding this and a fix?
- and last but not least, am I doing something wrong?
Thank you for your help.

There were some Macbook/SMB specific problems that suppossedly got fixed with the 10.4.6 update, but I'd try the 10.4.7 Combo Update, as Deltas have more problems.

BPM fork problem

Similar Messages

Maybe you are looking for