Rollback Issue in BPEL

Hi All,
I am facing issue with Txn rollback in BPEL while using fault policies.My BPEL polls JMS Q and invokes 2 databases and 1 jms adapter in a sequential manner. I want if any invocation fails , the preceding successful invocations should be rollbacked and txn should be retried as per the source jms q retry settings .I have worked on it using 2 approaches-
1. Without using Fault Policy- I created a catchall and added a throw activity with rollback fault in it. So if any invocation fails, the rollback and retry is working fine as per the source Q retry settings.
2.With Fault Policy- I created a fault policy and fault binding(composite/reference level).Also I created a catchall and added a throw activity with rollback fault in it.When any invocation fails, the fault policy is called, and the Ora:retry action executes and retry calling the failed adapter as per the retry configuration set in the policy file. If after all retries, still invocation fails, then transaction goes to CatchAll block.Here is the problem.The transaction is not getting rollbacked and does not even retry as per the JMS Q retry configuration.
My agenda is to have a process, where in if any invocation fails, that should be retried as per fault policy retry configuration.If still invocation is faulted after all retries, the preceding invocations should be rollbacked and the message should be rollbacked to source jms Q. Again bpel should be instantiated with new transaction. Each instance should itself retry as per fault policy retry configuration.It should keep on retrying(creating new instances) till source Q jms retry limit is achieved.
Please help..
Thanks
Rishu

Hi Ananth,
Thanks for your reply.
I am using fault policies for Error handling in all my interfaces instead of handling the faults using the catch block.So the same approach I need to follow in this interface as well , where I am facing the issue. In my affected interface, I have 3 invocations in sequence manner and want that If any invocation fails, the earlier successful transactions should be rollbacked.So when I am using fault policy, the failed invocation is getting retried as per the fault policy retry , but it is not rollbacking the earlier 2 successful invocations, which it should do to have an effective error handling. Same is getting accomplished by using Catch and throwing rollback error from there. I am ok even if Q retry should not get executed but I need to rollback the previous successfully executed invocations.Could you please let me know how to handle this scenario using fault policy..May be I am skipping something then..
Thanks and Regards,
Rishu Sharma

Similar Messages

Transaction Rollback Issue in SOA Suite

Hi All,
We have been facing the transaction rollback issue very frequently and looking for a concrete solution on this. We have a Asynchronous BPEL process which calls a webservice in a loop of say 20 iteration. In the loop sometimes we face this transaction rollback issue and apparently the CatchAll block doesn't capture this.
In many blogs and oracle metalink we have seen few settings for this and made as per below :
1. JTA Transaction timeout increase : we have it 10000 sec.
2. The Data Source settings : 2.1 : XA transaction timeout check-box selected
2.2 : The Initial Capacity : 10, Maximum Capacity : 200 and Capacity Increment = 5
3. soa-infra -> EJBs -> set the transaction timeout to 1200 for many of the recommended BPELs
4. DISTRIBUTED_LOCK_TIMEOUT = 120 sec.
In short, all the recommended settings are in place. But still, we are facing this issue. Does anyone know if anything is missing in the setting ? Or any other solution to the issue?
Any help is truly appreciated.
Thanks,
Ashwini

I think, even if you change the setting you may not get rid of this problem permanently. Try to avoid call to the webservice in a loop and look for any other alternate solution like JMS (thread should not block for the response of the webservice).

TopLink issue on Bpel 10.1.3.3

We are having an issue with the bpel flow on Bpel 10.1.3.3 where when it is inserting into the database, not all of the fields are saving to the database.Several of the fields are being stored as null in the database although there are values
that should be saved. I deployed the flow on both Bpel 10.1.2 and Bpel 10.1.3.3 which have same database connection. Bpel 10.1.2 has no problem. My flow is created by Bpel 10.1.2 designer. I also use JDveloper 10.1.3.3 to migrate the flow and then deploy it on Bpel 10.1.3.3. However, I am still facing an issue on Bpel 10.1.3.3.
The data being inserted to database which I saw it from Bpel console.
<description>CAR-47511-000073*EBS.01088</description>
<exchangeRateType>ALEXTEST</exchangeRateType>
<invoiceAmount>4086.82</invoiceAmount>
<invoiceCurrencyCode>USD</invoiceCurrencyCode>
<invoiceDate>2006-03-20T08:00:00</invoiceDate>
<invoiceId/>
<invoiceNum>ALEX-000002|EBS.000002</invoiceNum>
<orgId>204</orgId>
<invoiceTypeLookupCode>STANDARD</invoiceTypeLookupCode>
<source>TRANSPORTATION MANAGEMENT</source>
<vendorId>3932</vendorId>
<vendorSiteId>4646</vendorSiteId>
However, only invoice id, source, vendor id, vendor site id are inserted into dabase. All others are nulls.
Any idea? thanks !!

To complement information:
1/ Zoom in doesn't have this issue:
both Ctrl + Numpad + and Ctrl + = are working
On AZERTY keyboard = and + are on same key.
2/ When using Ctrl + Numpad - , I observe same behaviour as Ctrl + 6 (opens a box to comment)...
Can you at least fix Ctrl + Numpad - for zoom out?
(this won't be unversal but will at leas save 90% of the issues)

Performance issues in BPEL

We are having an integration wherein the messages coming via b2b are being passed on to the bpel process for further processing..
When we are testing with about 500 transactions means 500 separate files each having one transaction in it ..one ST SE transaction data block in it only about 10 transactions are being processed in parallel by the bpel process...while the remaining transactions go into the manual recovery queue.
so at a time only about 10 bpel instances are getting created and then once few of these transactions get completely processed by the bpel process..slowly the transactions from the Manual Recovery Queue gets assigned to a new bpel instance..means a new bpel instances get created..
we are expecting a load of about 16000 transactions for the entire day and the concern here is that we need all of these b2b transactions to be processed by the bpel process in about 5 to 6 hours...is it possible in any way that we change the configuration setting such that more number of parallel bpel instances could be trigerred at the same time so that this could increase the processing speed on the bpel side.
Could someone please help us out with this slowness issue in bpel.
thanks

Have a look at this doc
http://www.oracle.com/technology/tech/soa/soa-suite-best-practices/soa_best_practices_1013x_drop1.pdf
The section on threading should answer you question. You need to set the dspMaxThreads property to the appropraite level for your environment. Make sure your hardware is able to cope.
cheers
James

Investigate rollback issues

Hi,
How do I get more info on "Perform rollback" issues?
From SM21, I see an ID had "Perform rollback" issue. Double clicking on this entry will display info such as :
"system log message R6 8 : An error has caused an SAP rollback. All database updates will now be reset."
From the trace file (SM50) dev_w0, it shows :
M
M Tue Apr 14 08:30:37 2009
M ***LOG R68=> ThIRollBack, roll back () [thxxhead.c 13168]
There is nothing captured in Oracle alertlog and no short dumps (ST22).
How do I investigate such problems?
Thanks

Hi
Check this forum discussion [Link|http://sap.ittoolbox.com/groups/technical-functional/sap-basis/perform-rollback-233123]
Regards
Uday

DB Adapter rollback issue

Hi All,
We have a DB Adapter in our BPEL message flow to insert records to target DB and in case of error while invoking DB Adapter we are catching the error in catch and catch all block.
But in some scenarios like when we pass value of larger size to columns than what is declared in table( eg: if we pass 'YES' to column which datatype and size is varchar2(1)) then error is not caught in catch or catch all block, instead the process is getting rolled back with below exception.
We have tried with both XA connection and also non-XA connection but still the same issue. We have also set the JTA timeout to "2400". Could you please help us on this ?
Thanks in advance
<exception class="com.collaxa.cube.engine.EngineException">
JTA transaction is not in active state. The transaction became inactive when executing activity "" for instance "460,011", bpel engine can not proceed further without an active transaction. please debug the invoked subsystem on why the transaction is not in active status. the transaction status is "MARKED_ROLLBACK". The reason was The execution of this instance "460011" for process "MSITesseractDBInsert" is supposed to be in an active jta transaction, the current transaction status is "MARKED_ROLLBACK", the underlying exception is "BINDING.JCA-12563 Exception occured when binding was invoked. Exception occured during invocation of JCA binding: "JCA Binding execute of Reference operation 'insert' failed due to: DBWriteInteractionSpec Execute Failed Exception. insert failed. Descriptor name: [InsertCustomerDataToTesseractDB.MotCustomerLoadIn]. Caused by java.sql.BatchUpdateException: String or binary data would be truncated.. Please see the logs for the full DBAdapter logging output prior to this exception. This exception is considered not retriable, likely due to a modelling mistake. To classify it as retriable instead add property nonRetriableErrorCodes with value "-8152" to your deployment descriptor (i.e. weblogic-ra.xml). To auto retry a retriable fault set these composite.xml properties for this invoke: jca.retry.interval, jca.retry.count, and jca.retry.backoff. All properties are integers. ". The invoked JCA adapter raised a resource exception. Please examine the above error message carefully to determine a resolution. " . Consult the system administrator regarding this error.
<stack>
<f>com.oracle.bpel.client.util.TransactionUtils.throwExceptionIfTxnNotActive#107</f>
<f>com.collaxa.cube.ws.WSInvocationManager.invoke#352</f>
<f>com.collaxa.cube.engine.ext.common.InvokeHandler.__invoke#1070</f>
<f>com.collaxa.cube.engine.ext.common.InvokeHandler.handleNormalInvoke#584</f>
<f>com.collaxa.cube.engine.ext.common.InvokeHandler.handle#132</f>
<f>com.collaxa.cube.engine.ext.bpel.common.wmp.BPELInvokeWMP.__executeStatements#74</f>
<f>com.collaxa.cube.engine.ext.bpel.common.wmp.BaseBPELActivityWMP.perform#166</f>
<f>com.collaxa.cube.engine.CubeEngine.performActivity#2687</f>
<f>com.collaxa.cube.engine.CubeEngine._handleWorkItem#1190</f>
<f>com.collaxa.cube.engine.CubeEngine.handleWorkItem#1093</f>
<f>com.collaxa.cube.engine.dispatch.message.instance.PerformMessageHandler.handleLocal#76</f>
<f>com.collaxa.cube.engine.dispatch.DispatchHelper.handleLocalMessage#218</f>
<f>com.collaxa.cube.engine.dispatch.DispatchHelper.sendMemory#297</f>
<f>com.collaxa.cube.engine.CubeEngine.endRequest#4609</f>
<f>com.collaxa.cube.engine.CubeEngine.endRequest#4540</f>
<f>com.collaxa.cube.engine.CubeEngine._createAndInvoke#713</f>
<f>...</f>
</stack>Regards,
Ramesh
Edited by: Ramesh Menon on Mar 7, 2013 1:12 PM

Hi All,
We have a DB Adapter in our BPEL message flow to insert records to target DB and in case of error while invoking DB Adapter we are catching the error in catch and catch all block.
But in some scenarios like when we pass value of larger size to columns than what is declared in table( eg: if we pass 'YES' to column which datatype and size is varchar2(1)) then error is not caught in catch or catch all block, instead the process is getting rolled back with below exception.
We have tried with both XA connection and also non-XA connection but still the same issue. We have also set the JTA timeout to "2400". Could you please help us on this ?
Thanks in advance
<exception class="com.collaxa.cube.engine.EngineException">
JTA transaction is not in active state. The transaction became inactive when executing activity "" for instance "460,011", bpel engine can not proceed further without an active transaction. please debug the invoked subsystem on why the transaction is not in active status. the transaction status is "MARKED_ROLLBACK". The reason was The execution of this instance "460011" for process "MSITesseractDBInsert" is supposed to be in an active jta transaction, the current transaction status is "MARKED_ROLLBACK", the underlying exception is "BINDING.JCA-12563 Exception occured when binding was invoked. Exception occured during invocation of JCA binding: "JCA Binding execute of Reference operation 'insert' failed due to: DBWriteInteractionSpec Execute Failed Exception. insert failed. Descriptor name: [InsertCustomerDataToTesseractDB.MotCustomerLoadIn]. Caused by java.sql.BatchUpdateException: String or binary data would be truncated.. Please see the logs for the full DBAdapter logging output prior to this exception. This exception is considered not retriable, likely due to a modelling mistake. To classify it as retriable instead add property nonRetriableErrorCodes with value "-8152" to your deployment descriptor (i.e. weblogic-ra.xml). To auto retry a retriable fault set these composite.xml properties for this invoke: jca.retry.interval, jca.retry.count, and jca.retry.backoff. All properties are integers. ". The invoked JCA adapter raised a resource exception. Please examine the above error message carefully to determine a resolution. " . Consult the system administrator regarding this error.
<stack>
<f>com.oracle.bpel.client.util.TransactionUtils.throwExceptionIfTxnNotActive#107</f>
<f>com.collaxa.cube.ws.WSInvocationManager.invoke#352</f>
<f>com.collaxa.cube.engine.ext.common.InvokeHandler.__invoke#1070</f>
<f>com.collaxa.cube.engine.ext.common.InvokeHandler.handleNormalInvoke#584</f>
<f>com.collaxa.cube.engine.ext.common.InvokeHandler.handle#132</f>
<f>com.collaxa.cube.engine.ext.bpel.common.wmp.BPELInvokeWMP.__executeStatements#74</f>
<f>com.collaxa.cube.engine.ext.bpel.common.wmp.BaseBPELActivityWMP.perform#166</f>
<f>com.collaxa.cube.engine.CubeEngine.performActivity#2687</f>
<f>com.collaxa.cube.engine.CubeEngine._handleWorkItem#1190</f>
<f>com.collaxa.cube.engine.CubeEngine.handleWorkItem#1093</f>
<f>com.collaxa.cube.engine.dispatch.message.instance.PerformMessageHandler.handleLocal#76</f>
<f>com.collaxa.cube.engine.dispatch.DispatchHelper.handleLocalMessage#218</f>
<f>com.collaxa.cube.engine.dispatch.DispatchHelper.sendMemory#297</f>
<f>com.collaxa.cube.engine.CubeEngine.endRequest#4609</f>
<f>com.collaxa.cube.engine.CubeEngine.endRequest#4540</f>
<f>com.collaxa.cube.engine.CubeEngine._createAndInvoke#713</f>
<f>...</f>
</stack>Regards,
Ramesh
Edited by: Ramesh Menon on Mar 7, 2013 1:12 PM

Unique rollback in a BPEL process using web services

We are a cellular communication company using Oracle DB (10g) and Oracle Forms to run our business applications (customer care and the like). We have planned a system migration using the SOA approach with the latest release of Oracle BPEL Suite (business logic), Oracle ADF BC (presentation level), Oracle OAS (application server) and JDeveloper (application development IDE).
Within our current applications we have processes that perform several tasks (inserts, updates and/or deletes on the same DB) as part of one atomic DB transaction. Using the SOA approach we’ve deemed those tasks as being ‘services’ that we, let’s say for the sake of argument, could implement either as web services (so as to be published for reuse) or as stored procedures encapsulated via DB adapters (BPEL lexicon).
Our biggest challenge, and hence our question to the SOA world, is how to attain that same unique transaction. We understand that there is no one perfect approach, for the ‘solution’ depends on too many factors and choices; what we’d like to know are past experiences people had or, at best, what do people consider what the best practices in the industry are regarding this matter.
Approach A: all services are implemented as DB adapters (stored procedures and the like).
PROS: one single DB rollback should any service fail to execute properly and so undo all changes made so far.
CONS: highly coupled, non-visible components (via UDDI); impossibility to manage, control and audit (via WS Manager) components; weak/poor SOA Governance.
Approach B: all services are implemented as web services.
PROS: loosely coupled, self-contained, and web accessible visible components (via UDDI) besides the ease to manage, control and audit (via WS Manager); strong SOA Governance.
CONS: since each web service is a ‘different’ transaction with its own ‘commit’, a unique rollback transaction for the whole process will not be possible should a WS fail, for previous WSs will have already committed their data changes before doing the rollback.
Approach C: services implemented as a mix of web services and DB Adapters.
PROS: those of Approach B for web services.
CONS: those of Approach A (DB adapters) + those of Approach B (web services).
<hr />
So, the 64,000 dollar question is how to attain a single transaction when you have web services as part of your process? Do we need to ’build/create’ a schema/functionality similar to the ‘DB-undo-segment’ in order to obtain a single rollback transaction for the whole process or does the BPEL COMPENSATE delivers this functionality somehow?
Are we not seeing the obvious? Are we not seeing the forest because we’re looking at a tree? Please keep in mind that we are using Oracle BPEL.
Thanks a bunch,
--Sergio

Hi Marinus,
you are absolutely right and not missing anything. OWSM does not free you from taking additional measures to block direct interaction with the unsecure WebService from not "authorized" clients. You must restrict the http response to the OWSM host only (either in the Apache or the OC4J configuration). See the related documentation.
--olaf

Java.lang.RuntimeException: failed to compile execlets (JAVA embeding issue in BPEL 2.0)

HI i was trying to use java code in BPEL 2.0 by using java Embedding. while deploying the application getting the error.
even after importing the corrsponding class
below is the code snippet for the same :
<import location="oracle.soa.management.facade.Locator"
 importType="http://schemas.oracle.com/bpel/extension/java"/>
 <import location="oracle.soa.management.facade.LocatorFactory"
 importType="http://schemas.oracle.com/bpel/extension/java"/>
 <import location="oracle.soa.management.facade.Component"
 importType="http://schemas.oracle.com/bpel/extension/java"/>
<extensionActivity>
 <bpelx:exec name="Java_Embedding1" language="java">
 <![CDATA[Locator locator = LocatorFactory.createLocator();
String compositeRevision = (String)getVariableData("compositeRevision");
String compositeNameRevision = "ASG/ImportCustomerData!"+ compositeRevision;
Composite prowessCustomerComposite = (Composite)locator.lookupComposite(compositeNameRevision);
CompositeInstanceFilter filter = new CompositeInstanceFilter();
filter.setState(CompositeInstance.STATE_RUNNING);]]>
 </bpelx:exec>
 </extensionActivity>
while deploying gettting below error.
06:18:58 PM] Error deploying archive sca_ImportCutomerData_rev1.0.jar to partition "default" on server AdminServer [http://LTRTHIPIREDDY.keste.com:7001]
[06:18:58 PM] HTTP error code returned [500]
[06:18:58 PM] Error message from server:
There was an error deploying the composite on AdminServer: Deployment Failed: Error occurred during deployment of component: importCustomerData_BPEL to service engine: implementation.bpel, for composite: ImportCutomerData: ORABPEL-05250
Error deploying BPEL suitcase.
error while attempting to deploy the BPEL component file "E:\Oracle\Middleware\user_projects\domains\Raja_domain\servers\AdminServer\dc\soa_3ca99378-812f-460f-9760-d19a15c0ef44"; the exception reported is: java.lang.RuntimeException: failed to compile execlets of importCustomerData_BPEL
This error contained an exception thrown by the underlying deployment module.
Verify the exception trace in the log (with logging level set to debug mode).
[06:18:58 PM] Check server log for more details.
[06:18:58 PM] Error deploying archive sca_ImportCutomerData_rev1.0.jar to partition "default" on server AdminServer [http://LTRTHIPIREDDY.keste.com:7001]
[06:18:58 PM] #### Deployment incomplete. ####
[06:18:58 PM] Error deploying archive file:/C:/JDeveloper/mywork_Arbonne/CustomerInboundApp/ImportCutomerData/deploy/sca_ImportCutomerData_rev1.0.jar
(oracle.tip.tools.ide.fabric.deploy.common.SOARemoteDeployer)
Thanks in Advance...
rajashekar

Hi have solved the issue by importing the class file.
<import location="oracle.soa.management.facade.Locator"
 importType="http://schemas.oracle.com/bpel/extension/java"/>
<import location="oracle.soa.management.facade.LocatorFactory"
 importType="http://schemas.oracle.com/bpel/extension/java"/>
<import location="oracle.soa.management.facade.Composite"
 importType="http://schemas.oracle.com/bpel/extension/java"/>
<import location="java.util.List"
 importType="http://schemas.oracle.com/bpel/extension/java"/>
<import location="oracle.soa.management.util.CompositeInstanceFilter"
 importType="http://schemas.oracle.com/bpel/extension/java"/>
<import location="oracle.soa.management.facade.CompositeInstance"
 importType="http://schemas.oracle.com/bpel/extension/java"/>

DB Transaction ROLLBACK issues

Hi All,
I am running BPEL version 10.1.2.0.2, and i am trying out DB Transactional Insert. It looks like it is not doing proper rollback when an exception is occuring in BPEL process flow as well as the audit trace is also not generated when an exception happens. Is this a bug in this version and is there a patch available.
I have set the property in bpel.xml as transaction=participate and in partner link participate=true. For Transactionality is there any other setting to be done in any config xml files like /../../data-source.xml or oc4j-ra.xml
Appreciate if any one can let me know if i need to do any other settings
Thank you
RM

According to Clemens at his note:
there are some consideration for making this happen ..
a) process type should be sync (to ensure the process runs in one thread)
b) on each partnerlink specify participate=true (as property) and
c) db adapter's database connection MUST point to a XA-location to participate in the global tx ..
Re: the transaction on multiple PL/SQL Web Service on once BPEL process

Photoshop Elements 9 'Rollback Issue'?

Everytime I try to install Photostop Elements 9, it finishes installing, and just when it starts registering the product, it fails and has to rollback, and the program is uninstalled. Im running Windows 7 Home Premium 64 Bit.
Can anyone help or suggest something I do?

OMG...I have been dealing with this issue ALL day and it is now 11:51 pm!!! I had purchased a new, sealed, PSE 9 on dvd, tried to install it...almost got to the end and it started rolling back. got the same screen about the Installshield. Spent hours researching a fix on the internet...nothing worked. Finally found and let an Adobe tech take control of my computer. He tried a couple of different things but the rollback continued. Then he tried to make another administrator account (I'm on XP media edition) and he hit restart...and was gone. Then I had to spend a couple hours trying to get my settings and configurations back to the way they were (for some reason the computer reverted to the programs and setting it originally came with and system restore wouldn't work)
Finally got things back to how they were, went on to Adobe and purchased a download copy of PSE 9 from their store. More money again but I thought if it works I really don't care. All excited and spent the time downloading the program, went to install it, and same darn rolling back problem. Tried a couple more times in disbelief...no change.
Then did research into more Installshield problems and possible fixes and then updates, patches, drivers and finally researched reinstall. Found it finally on c-net, downloaded and reinstalled Installshield (it copied, replaced, fixed, cleaned etc as I watched). Excitedly I tried to install the downloaded copy of PSE 9...but the rolling back continues!!!!!!!!
I have NEVER had this problem with installing software before (downloaded or hard copies). Please please please someone figure this out!!!!!!!!!!!!!!
Thanks,
Linda

Ref:Siebel webservice calling issue in Bpel

Hi,
Hope you are doing good. Kindly share you valuable ideas regards below issue please.
I have developed one mediator with 4 Operations namely 1). InsertOrUpdateContactSSM Operation
2).UpdateContact Operation
3). UpdateContact Operations
4).QueryByIdContact Operation.
I have created 4 BPEL process's for above 4 Operations. Here my SOA suite version is 11.1.1.6.
I have invoking one Siebel webservice, In this webservice having above operation defined.
It's developed, it's complied without error and i have deployed into my SOA server it's deployed successfully.
At testing time it's throwing below :
Exception occured when binding was invoked. Exception occured during invocation of JCA binding: "JCA Binding execute of Reference operation 'InsertOrUpdateContactSSM' failed due to: *SOAP-ENV:Server, error code: There is no active Web Service with operation named 'http://www.absa.co.za/siebel/webservices:InsertOrUpdateContactSSM_Input'.(SBL-EAI-04313)".* The invoked JCA adapter raised a resource exception. Please examine the above error message carefully to determine a resolution.
Any one can give proved solution for above error. It's bit Urgent.
Thanks in Advance.
regards
Anil

Follow the instructions mentioned at this blog to resolve "SBL-EAI-04313" Error
http://www.soastation.org/2007/05/troubleshooting-siebel-web-services-no.html
Here is the another link with screen shots:
https://blogs.oracle.com/pt/entry/siebel_web_service_invocation
Mark the Posting appropriately as "helpful" or "Correct Answer", if your issue is solved.
Edited by: 980644 on Feb 27, 2013 6:10 PM

Issue calling BPEL process in other domains

I just moved my code from the development environment to test and I found an issue where the BPEL processes can't call one another if they reside in a different domain. There's no issue in calling a BPEL process from another BPEL process if they are in the same domain, but it just doesn't work if they are in different domains. The odd thing is that I'm not getting any errors either. I looked at the BPEL admin and everything seems to be consistent with what's in the development environment where everything works, so is there somewhere else I should look? thanks in advance.

Are your domain names in lower case? There are several problems when using domain names which contain uppercase letters.
Andre

Rollback with XATransaction bpel 2.0 pl/sql

Hello,
I've added in the composite.xml:
<component name="Selector" version="2.0">
<implementation.bpel src="Selector.bpel"/>
<property name="bpel.config.transaction" many="false" type="xs:string">required</property>
<property name="partnerLink.DB_INSERTAR_LOG.idempotent" type="xs:string"
many="false">false</property>
<property name="partnerLink.DB_ALMACENAR_MENSAJE.idempotent"
type="xs:string" many="false">false</property>
</component>
and also I've added:
<throw name="Throw1" faultName="bpelx:rollback"/>
The database Adapter invoke a procedure pl/sql.
Could anybody help me??
Thanks

you can execute sql statment for Rollback as
begin
rollback
end;
This will rollback the transaction

WebLogicJtaTransactionManager rollback issue

Hi All,
I have some problem using WebLogicJtaTransactionManager. My problem is although there is an exception, the transaction cannot rollback.
I'm using hibernate + spring and weblogic as the application server.
this is my code:
<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://www.springframework.org/schema/beans"
 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
 xmlns:tx="http://www.springframework.org/schema/tx"
 xmlns:aop="http://www.springframework.org/schema/aop"
 xmlns:jee="http://www.springframework.org/schema/jee"
 xsi:schemaLocation="
 http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-2.0.xsd
 http://www.springframework.org/schema/tx http://www.springframework.org/schema/tx/spring-tx-2.0.xsd
 http://www.springframework.org/schema/aop http://www.springframework.org/schema/aop/spring-aop-2.0.xsd
 http://www.springframework.org/schema/jee http://www.springframework.org/schema/jee/spring-jee.xsd">
 <jee:jndi-lookup id="dataSource" jndi-name="localDS" />
 <jee:jndi-lookup id="myNetsDataSource" jndi-name="myNetsDS" />
 <bean id="sessionFactory"
 class="org.springframework.orm.hibernate3.LocalSessionFactoryBean" destroy-method="destroy">
 <property name="dataSource" ref="dataSource" />
 <property name="configurationClass">
 <value>org.hibernate.cfg.AnnotationConfiguration</value>
 </property>
 <property name="configLocations">
 <list>
 <value>/WEB-INF/hibernate.cfg.xml</value>
 </list>
 </property>
 <property name="hibernateProperties">
 <props>
 <prop key="hibernate.connection.driver_class">${hibernate.connection.driver_class}</prop>
 <prop key="hibernate.dialect">${hibernate.dialect}</prop>
 <prop key="hibernate.current_session_context_class">${hibernate.current_session_context_class}</prop>
 <prop key="hibernate.show_sql">${hibernate.show_sql}</prop>
 <prop key="hibernate.format_sql">${hibernate.format_sql}</prop>
 </props>
 </property>
 </bean>
 <bean id="myNetsSessionFactory"
 class="org.springframework.orm.hibernate3.LocalSessionFactoryBean" destroy-method="destroy">
 <property name="dataSource" ref="myNetsDataSource" />
 <property name="configurationClass">
 <value>org.hibernate.cfg.AnnotationConfiguration</value>
 </property>
 <property name="configLocations">
 <list>
 <value>/WEB-INF/myNets-hibernate.cfg.xml</value>
 </list>
 </property>
 <property name="hibernateProperties">
 <props>
 <prop key="hibernate.connection.driver_class">${hibernate.connection.driver_class}</prop>
 <prop key="hibernate.dialect">${hibernate.dialect}</prop>
 <prop key="hibernate.current_session_context_class">${hibernate.current_session_context_class}</prop>
 <prop key="hibernate.show_sql">${hibernate.show_sql}</prop>
 <prop key="hibernate.format_sql">${hibernate.format_sql}</prop>
 </props>
 </property>
 </bean>
 <bean id="txManager"
 class="org.springframework.transaction.jta.WebLogicJtaTransactionManager">
 <property name="transactionManagerName"
value="javax.transaction.TransactionManager"/>
 </bean>
 <aop:config>
 <aop:pointcut
 expression="execution(* com.kiosk.commonlib.service.*.*(..))"
 id="ats" />
 <aop:advisor advice-ref="txAdvice" pointcut-ref="ats" />
 </aop:config>
 <tx:advice id="txAdvice" transaction-manager="txManager">
 <tx:attributes>
 <tx:method name="save*" propagation="REQUIRED" />
 <tx:method name="delete*" propagation="REQUIRED" />
 <tx:method name="remove*" propagation="REQUIRED" />
 <tx:method name="update*" propagation="REQUIRED" />
 <tx:method name="log*" propagation="REQUIRED" />
 <tx:method name="get*" propagation="REQUIRED"
 read-only="true" />
 <tx:method name="list*" propagation="REQUIRED"
 read-only="true" />
 </tx:attributes>
 </tx:advice>
</beans>
I created two datasource (localDS and myNetsDS) in weblogic server, both access to different database. is there something wrong with my txmanager configuration?
Thanks

Hi Ananth,
Thanks for your reply.
I am using fault policies for Error handling in all my interfaces instead of handling the faults using the catch block.So the same approach I need to follow in this interface as well , where I am facing the issue. In my affected interface, I have 3 invocations in sequence manner and want that If any invocation fails, the earlier successful transactions should be rollbacked.So when I am using fault policy, the failed invocation is getting retried as per the fault policy retry , but it is not rollbacking the earlier 2 successful invocations, which it should do to have an effective error handling. Same is getting accomplished by using Catch and throwing rollback error from there. I am ok even if Q retry should not get executed but I need to rollback the previous successfully executed invocations.Could you please let me know how to handle this scenario using fault policy..May be I am skipping something then..
Thanks and Regards,
Rishu Sharma

An issue with BPEL fault handling

Hi,
I need to catch an exception in my BPEL process. The exception is thrown from a partner service(in java). I am using the catch fault handler in BPEL to catch a specific fault.
1. When I give "systemFault" as the faultName its catching the fault
2. But when I give "Exception" as the faultName its not catching the fault
In the second case server log gives me the following fault,
BPCOR-6135:A fault was not handled in the process scope; Fault Name is {......
Can you please help me to figure out the problem here.
Please let me know if you need any more information.
Thanks in advance,
Jose John

Thank you very much for your reply.
Following is the wsdl file,
relevant part of the WSDL file
<message name="Exception">
<part name="fault" element="tns:Exception"/>
</message>
<portType name="PostConstraintChecker">
<operation name="check">
<input message="tns:check"/>
<output message="tns:checkResponse"/>
<fault message="tns:Exception" name="Exception"/>
</operation>
</portType>
<binding name="PostConstraintCheckerPortBinding" type="tns:PostConstraintChecker">
<soap:binding transport="http://schemas.xmlsoap.org/soap/http" style="document"/>
<operation name="check">
<soap:operation soapAction=""/>
<input>
<soap:body use="literal"/>
</input>
<output>
<soap:body use="literal"/>
</output>
<fault name="Exception">
<soap:fault name="Exception" use="literal"/>
</fault>
</operation>
</binding>
Is this correct? Do you see any problems in this? Please let me know if you need any more information which will help you in debugging the issue.
Thanks,
Jose John

Rollback Issue in BPEL

Similar Messages

Maybe you are looking for