Merge BPEL Servers

Hi,
We have an interesting issue. We have a legacy 10.1.3.1 instance that sits on older kit and runs a single application. We have a new 10.1.3.4 HA Cluster implementation that is running a number of new BPEL Applications. The legacy BPEL Processes are long running and there is no natural point to end processing. We need to port the legacy system to the new one. We cannot just stop instance, export hydration store, import to new system...etc as the new system is already in use.
As I see it, we have a couple of options:
1. Move code to new system and run in parallel. All new processes kicked off on new kit and legacy processes allowed to complete on old kit. Gets a little tricky as we use Human Task Work Lists and have hook in's from ADF. Can get it to work, but very complicated.
2. Attempt to add the hydration store metadata to the new hydration store. This would be extremely complex I am guessing as we would need to resolve any Primary and Foreign Key conflicts between old system data and the data already stored in the new hydration store.
3. As all BPEL processes on new kit are all fast to complete, we could pick a point on new kit to prevent any new processes from being kicked off. Purge the hydration stores completely (including undeploying all code). Migrate the legacy system and then re-deploy all existing new code. Downside is we lose our existing history on new kit for new processes.
4. Any other suggestions welcome!!
Thanks
Iain

Yes and no. The BPEL process itself is based on the BPEL standard and can be used on any other server. But the BPEL process is packed into a SuiteCase, jar file. This file contains other files like bpel.xml, these is vendor specific. So you can not run a BPEL process designed in JDev into another non Oracle BPEL server, without any changes.

Similar Messages

JDeveloper BPEL Designer with other BPEL Servers

Can the BPEL XML emitted from JDeveloper designed be successfully executed by servers other than Oracle BPEL Process Manager?

Yes and no. The BPEL process itself is based on the BPEL standard and can be used on any other server. But the BPEL process is packed into a SuiteCase, jar file. This file contains other files like bpel.xml, these is vendor specific. So you can not run a BPEL process designed in JDev into another non Oracle BPEL server, without any changes.

High Availability of BPEL System

We are having a High Availability architecture configuration for the BPEL System in our production environment.
The BPEL servers are clustered in the middle tier of the architecture and RAC is used in the database tier of the architecture.
We have 5 BPEL processes which are getting invoked within each other. For eg:
BPELProcess1 --> BPELProcess2 --> BPELProcess3, BPELProcess4 &
BPELProcess4 --> BPELProcess5
Now when all the above BPEL processes are deployed on both the nodes of the BPEL server, how do we handle the end point URL's of these BPEL servers.
Should we hardcode the end point URL in the invoking BPEL process or should we replace the IP address of the two nodes of the BPEL server with the IP address of the load balancer.
If we replace the IP address of the BPEL server with the IP address of the load balancer, it will require us to modify, redeploy and retest all the BPEL processes again.
Please advise
Thanks

The BPEL servers are configured with active - active topology and RAC is used in the database tier of the architecture.
BPEL Servers is not clustered. Load Balancer is used in front of the two nodes of the BPEL servers.

RAC node outage causes SOA Suite 10.1.3.4 BPEL failure

Using weblogic 9.2 and the SOA Suite 10.1.3.4. We use a 10g Oracle RAC ( 2 nodes ); the WL cluster has a multi data source of 2 pools, each pool pointing to a single node in the rac, each pool deployed to the cluster, and the multi data source in load-balancing mode.
So the other night, one of the db nodes had a hardware failure ( ironically, with a remote monitoring / management card ). Annoying, but it should not have caused the BPEL servers to be in "FAILED NOT RESTARTABLE" status the next morning.
Jun 9, 2009 12:10:07 AM EDT> <Warning> <JDBC> <BEA-001129> <Received exception while creating connection for pool "esbaqds2": Io exception: The Network Adapter could not establish the connection>
SEVERE: Destroying JMSDequeuer failed
oracle.jms.AQjmsException: Connection has been administratively destroyed. Reconnect.
at oracle.jms.AQjmsSession.preClose(AQjmsSession.java:980)
at oracle.jms.AQjmsObject.close(AQjmsObject.java:409)
at oracle.jms.AQjmsSession.close(AQjmsSession.java:1020)
at oracle.tip.esb.server.dispatch.JMSDequeuer.destroy(JMSDequeuer.java:419)
at oracle.tip.esb.server.dispatch.JMSDequeuer.destroyWithoutUnsubscribing(JMSDequeuer.java:395)
at oracle.tip.esb.server.dispatch.JMSDequeuer.dequeue(JMSDequeuer.java:175)
at oracle.tip.esb.server.dispatch.agent.ESBWork.process(ESBWork.java:174)
at oracle.tip.esb.server.dispatch.agent.ESBWork.run(ESBWork.java:132)
at weblogic.connector.security.layer.WorkImpl.runIt(WorkImpl.java:108)
at weblogic.connector.security.layer.WorkImpl.run(WorkImpl.java:44)
at weblogic.connector.work.WorkRequest.run(WorkRequest.java:93)
at weblogic.work.ServerWorkManagerImpl$WorkAdapterImpl.run(ServerWorkManagerImpl.java:518)
at weblogic.work.ExecuteThread.execute(ExecuteThread.java:209)
at weblogic.work.ExecuteThread.run(ExecuteThread.java:181)
followed by a 2 GB log file containing 1.3 million iterations of the following within the next 10 minutes before the managed servers failed.
java.lang.NullPointerException
at java.lang.String.<init>(String.java:144)
at oracle.tip.esb.server.dispatch.JMSDequeuer.dequeue(JMSDequeuer.java:168)
at oracle.tip.esb.server.dispatch.agent.ESBWork.process(ESBWork.java:174)
at oracle.tip.esb.server.dispatch.agent.ESBWork.run(ESBWork.java:132)
at weblogic.connector.security.layer.WorkImpl.runIt(WorkImpl.java:108)
at weblogic.connector.security.layer.WorkImpl.run(WorkImpl.java:44)
at weblogic.connector.work.WorkRequest.run(WorkRequest.java:93)
at weblogic.work.ServerWorkManagerImpl$WorkAdapterImpl.run(ServerWorkManagerImpl.java:518)
at weblogic.work.ExecuteThread.execute(ExecuteThread.java:209)
at weblogic.work.ExecuteThread.run(ExecuteThread.java:181)
Both managed instances of the BPEL cluster failed, even though the 1st node of the Oracle RAC was still available.
Our 10.3 cluster, also using multi data sources to the same RAC for the OSB components, simply went on about its business using the remaining rac node pool.
Seems to be a single point of failure...

We haven't changed the JDBC connection string yet, but we did run a test in the same environment while Oracle support considers the situation.
For the test, we simply shutdown one node of the RAC and watched to see what happens. Within the space of a minute, the JDBC "Failed Reserve Request Count" was increasing by thousands on every refresh of the screen. We restarted the RAC node after 5 minutes, by which time the "Failed Reserve Request Count" was over 190,000
The 2 BPEL managed servers remained in Running status and each created a 660 MB log file within that 5 minutes. In the original outage, the nodes were down for about 15 minutes. Most of the logging is being generated from within the oracle.tip.esb classes, not by the weblogic classes. It looks like that once the pool pointing to the downed RAC node becomes disabled, the Oracle BPEL code is still trying to use it even though the multi-source JNDI is the published lookup:
INFO: JMSDequeuer::createConnection - AQ Topics
java.sql.SQLException: weblogic.common.ResourceException:
esbaqds(esbaqds2): Pool esbaqds2 is disabled, cannot allocate resources to applications..
esbaqds(esbaqds1): Pool esbaqds1 is disabled, cannot allocate resources to applications..
at weblogic.jdbc.common.internal.JDBCUtil.wrapAndThrowResourceException(JDBCUtil.java:250)
at weblogic.jdbc.common.internal.RmiDataSource.getPoolConnection(RmiDataSource.java:348)
at weblogic.jdbc.common.internal.RmiDataSource.getConnection(RmiDataSource.java:364)
at oracle.tip.esb.server.dispatch.JMSDequeuer.createAQConnection(JMSDequeuer.java:559)
at oracle.tip.esb.server.dispatch.JMSDequeuer.dequeue(JMSDequeuer.java:159)
at oracle.tip.esb.server.dispatch.agent.ESBWork.process(ESBWork.java:174)
at oracle.tip.esb.server.dispatch.agent.ESBWork.run(ESBWork.java:132)
at weblogic.connector.security.layer.WorkImpl.runIt(WorkImpl.java:108)
at weblogic.connector.security.layer.WorkImpl.run(WorkImpl.java:44)
at weblogic.connector.work.WorkRequest.run(WorkRequest.java:93)
at weblogic.work.ServerWorkManagerImpl$WorkAdapterImpl.run(ServerWorkManagerImpl.java:518)
at weblogic.work.ExecuteThread.execute(ExecuteThread.java:209)
at weblogic.work.ExecuteThread.run(ExecuteThread.java:181)
Caused by: weblogic.common.ResourceException:
esbaqds(esbaqds2): Pool esbaqds2 is disabled, cannot allocate resources to applications..
esbaqds(esbaqds1): Pool esbaqds1 is disabled, cannot allocate resources to applications..
at weblogic.jdbc.common.internal.MultiPool.searchLoadBalance(MultiPool.java:331)
at weblogic.jdbc.common.internal.MultiPool.findPool(MultiPool.java:202)
at weblogic.jdbc.common.internal.ConnectionPoolManager.reserve(ConnectionPoolManager.java:77)
at weblogic.jdbc.common.internal.RmiDataSource.getPoolConnection(RmiDataSource.java:346)
... 11 more
SEVERE: Failed to process deferred message
oracle.tip.esb.server.dispatch.QueueHandlerException: Error creating "weblogic.common.ResourceException: No good connections available."
at oracle.tip.esb.server.dispatch.JMSDequeuer.createAQConnection(JMSDequeuer.java:661)
at oracle.tip.esb.server.dispatch.JMSDequeuer.dequeue(JMSDequeuer.java:159)
at oracle.tip.esb.server.dispatch.agent.ESBWork.process(ESBWork.java:174)
at oracle.tip.esb.server.dispatch.agent.ESBWork.run(ESBWork.java:132)
at weblogic.connector.security.layer.WorkImpl.runIt(WorkImpl.java:108)
at weblogic.connector.security.layer.WorkImpl.run(WorkImpl.java:44)
at weblogic.connector.work.WorkRequest.run(WorkRequest.java:93)
at weblogic.work.ServerWorkManagerImpl$WorkAdapterImpl.run(ServerWorkManagerImpl.java:518)
at weblogic.work.ExecuteThread.execute(ExecuteThread.java:209)
at weblogic.work.ExecuteThread.run(ExecuteThread.java:181)

BPEL process initiated by a database adapter causing problems in HA environ

We are having a High Availability architecture configuration(active - active) for the BPEL System in our production environment.
The BPEL servers are clustered in the middle tier of the architecture and RAC is used in the database tier of the architecture.
We have a BPEL process which is initiated by a database adapter.
This BPEL process polls a source database, reads the data, writes it into a JMS Topic and marks flag as read back to the table.
A Message Driven Bean(MDB) listens to the JMS Topic, and writes the data in the destination database.
We have the polling process and the MDB deployed to both the nodes of the BPEL server.
We have noticed that some records are getting processed by both the nodes.
This is resulting in duplicate records at the destination tables.
Kindly advise how we can resolve this issue.

The BPEL servers are configured with active - active topology and RAC is used in the database tier of the architecture.
BPEL Servers is not clustered. Load Balancer is used in front of the two nodes of the BPEL servers.

Distributing BPEL processes

Hi,
Does anyone have experience distributing BPEL processes across machines. I essentially want to split my process onto two machines so as not to overload a single box. So the questions are:
1. would i need 2 bpel servers ?
2. if ans wer to 1 is yes then would I need to make web service calls between these processes running on different boxes ..
regards,
v

Vikky,
Are you using 10g?
If BPELConsole is up but you don't see any process, you can go to the log files and see what is the problem.
It might relates to permissions, or DB issues etc....
Arik

BPEL Processes are Hanging.......

Hi,
I am using BPEL 10.1.2.0.2 version. The logic am implementing in BPEL process is a frontend application will invoke BPEL process.BPEL will take some inputs from frontend and invokes a procedure in database using DbAdapter.BPEL will pull the output of the procedure and will pass the same to Frontend.
The logic being implemented in procedures is select and update depends on the requirement.
We completed our development and moved to UAT and everything is running fine.From past one week we are facing a problem that BPEL processes are hanging when they are invoked both from BPEL console and Frontend. The process neither displaying any output nor throwing any exception.
After bouncing the BPEL servers everything is coming to normal.I didn't found any error log in both domain.log and OraBPEL~OC4j_BPEL~default_island.log files.
Our BPEL Environment has 2 BPEL Mid Tiers connected via a loadbalancer
whether it could be related to cache size or datasource connections ????
Please suggest what could be the problem...........
Thanks
Chandra

Check the timeouts on the firewalls. Also look at Dead Connection Detection (DCD) on the database server. You will need to turn on your DCD to keep your JDBC connections alive.
Pete

BAM and BPEL

Hi,
Can we run both BPEL and BAM server on the same machine ?
I have installed BAM 10.1.3.1 on MS Windows XP,and i had previously installed SOA Suite 10.1.3 on my machine.
I want to connect BAM and BPEL through Jdeveloper sensor actions.
I am trying the OrderBooking tutorial.
When i try starting the SOA Suite i get error with opmn and the server is not coming up.
what should i do to bring up the SOA Suite when the BAM server is running.
Regards

Hi
I am trying to connect BAM and BPEL using BAM sensors.
The BAM and BPEL servers are running on different machines.
after deploying the BPEL process when i check the sensor tab in bpel console
,the action shows the null,and no sensor values are displayed even after creating an instance the BPEL process.
where am i gong wrong ?
from where can i get more docs on BAM sensors and conneecting BAM and BPEL through BAM sensors.
Regards
Ashwini

Some question on CORE BPEL ENGINE

Hello,
I've read Chapter 4 and now I've some question:
1-Clustering: What's you intend which expression "fail-over can be configured on the engine" and How? What Kind of clustering you realize (active-active, active-passive,...)?
2-I think that I don't exactly understand what you intend with "BPEL Compensating Transaction Model" could you give me more information about this?
thank you guys!
Andrea

1- I just completed a draft of a tech note on High Availability that I can share with you that explains the clustering model in detail. At a very high level, the BPEL Server is completely stateless and reliable, including a built-in recovery service that enables failover in case of failures. You don't need to cluster the app-server to be able to cluster BPEL servers (for either high availability or scalability).
2 - BPEL provides support for compensating transactions, which is an alternative transaction model when XA-style, ACID transactions can not be used (either due to the long-running nature of a "transaction" or the inclusion of services which don't support XA/JTA style transactions. We have code examples available for this and slides from our training course that discuss it.
Send me an email if you would like access to either of these documents. I'll also try to publish them to OTN in the next few days as tech notes, but in the meantime, I'm happy to share the information directly with any individuals who want it.
Dave
David Shaffer
Product Mgr, Oracle BPEL Process Manager
[email protected]
W: 650.506.1729
http://otn.oracle.com/bpel

BPEL Compliance

Hi,
Just wondering if the BPEL file generated is able to work with other BPEL/apps servers?
Regards,
Anthony

Yes. Both the designer and the server use native BPEL format as there internal format so you should be able to execute those files on other compliant BPEL servers or edit them using other BPEL compliant editors.
The only extension Oracle has make is the <java:exec> activity which allows you to embed when necessary Java code snippet within a BPEL process. If you are concerned with portability, you can avoid using exec and use instead the WSIF Java binding.
Edwin

BPEL redundancy

Hi!
we need to backup our bpel server, so in case of error on one server the operation load will be redirected to the second server. Maybe later they should be both running at the same to time when we do some load balnacing. Is ok the architecture, when there are two bpel servers above one Oracle DB with bpel schema? will that work without problems? what is the recommended architecture for this problem?
Thanks a lot for answer (or point me to the right documentation)
Tomas

Dan,
This should work without problem. On the management side, the only thing to keep in mind when you scope this out is exception management (at the file level, at the record level and at the db connectivity level). You need to think early on of a good retry strategy and potentially use human workflow or a dead letter queue when all retries have failed.
Please send up an email to collaxa_support_us is you have additional questions.
Good luck,
Edwin

BPEL cluster

Would the BPEL cluster have any performance improvement?
If I increase number of clusters from 1 to 5. Will I get 5 folds improvement in the BPEL servers performance?
I want to know more information about clusters, load balancer. Could anyone please give some pointers?
Thanks,
Ravi

Hi Ravi,
A cluster can improve reliability, throughput, and performance. Please refer -
http://blogs.oracle.com/reynolds/2009/06/clustering_soa_suite.html
http://blogs.oracle.com/reynolds/2009/10/what_i_learnt_about_clustering.html
http://blogs.oracle.com/reynolds/2010/05/building_a_soabpmbam_cluster_p.html
http://download.oracle.com/docs/cd/B31017_01/integrate.1013/b28980/clusteringsoa.htm (For 10g)
Regards,
Anuj

How many parallel BPEL child processes can be invoked on single BPEL server

Hi,
Can anyone pls let me know
1. How many parallel BPEL child processes can be invoked on single BPEL server at same time?
2. How number of Invoker threads, Worker threads and DB connection pools are affecting the BPEL server's listener thread spawning capacity?
Thanks in advance.
-S

Hi James
Here is my problem. I need to do this for load balancing. I have two instances of SOA Suite running on single machine, i.e. two different BPEL servers. Now i need 2 BPEL instances on each server, i.e. two SOA Suite and four BPEL instances (Two BPEL servers on each SOA Suite)
I tried to create a new OC4J instance on my app server and named it as OC4J_new. By default OC4J instance would have these application installed:
ascontrol
ccore
coreman
datatags
default
esb-dt
esb-rt
gateway
hw_services
javasso
orabpel
orainfra
policymanager
ruleauthor
rulehelp
But when i create new OC4J instance, i can see only these applications:
ascontrol
datatags
default
javasso
How do i get rest of the applications in my new OC4J Instance??
I tried deploying .EAR files but it failed all the time. What are possible fixes?
Many thanks in advance

BPEL database cleanup

Hi all,
In our production system we're experiencing an unacceptable growth in the orabpel tablespace. We are willing to keep little or no instance data. There is a stored procedure named delete_ci in the collaxa package, however I've noticed that this procedure does not delete the blob message data in the xml_document table. I've seen that the document_ci_ref table keeps ci's versus dockeys. But it seems like that's not enough, the xml_document table keeps more data than that (about invoke and delivery messages I guess). So, can anybody show me the way to delete all the information, everything related to completed instances in the BPEL database?
Regards,
SerkanC

Thanks a lot Marc, I've checked your blog and finally I found the purge_instances_oracle.sql script in metalink which creates a stored procedure named purge_instances. But now we have another problem, which is that the cleanup operation takes a very long time. Although we've setup a cron job to do this at 3 AM (when there is almost no load on the system) the database becomes non-responsive because of this script and we begin to get the following exception in bpel:
<2008-03-26 04:07:07,114> <ERROR> <default.collaxa.cube> <BaseCubeSessionBean::logError> Error while invoking bean "cube delivery": Cannot fetch a datasource connection.
The process domain was unable to establish a connection with the datasource with the connection URL "jdbc/BPELServerDataSource". The exception reported is: javax.resource.spi.ResourceAllocationException: Unable to allocate new connect
ion - pool is full and timed out waiting for a connection to be freed up.
Please check that the machine hosting the datasource is physically connected to the network. Otherwise, check that the datasource connection parameters (user/password) is currently valid.
Moreover, these connections do not timeout and we have to restart the bpel server in order to make the system usable. Another interesting thing is, this happens only one of our two bpel servers, the second server is not affected by this.
In patch number 6501312 there's a script named upgrade_6501312_oracle.sql which creates an index on the dockey column of the document_dlv_msg_ref table. I will try this and I'm hoping that it will solve this problem. But I would appreciate any comments about this problem, because if creating the index does not solve it, then I will be in deep trouble :)
Best regards,
SerkanC

BPEL Admin: Server / Log tab: Where this log is generated?

Hi All
There is a log configuration tab on BPEL Admin (where you can configure log for apache axis for instance), but I can´t locate where this log is beeing generated, can you help me please?
I have already checked log files on:
{OracleHome}/bpel/domains/xxx/log
{OracleHome}/j2ee/bpel/log
{OracleHome}/opmn/logs
Thanks

Hi Afonso
I´m trying to catch some log for Apache Axis or something related to SOAP messages, WSIF or WebServices call.
I have two BPEL process (lets say A and B).
Process A call process B using dynamic partner links (setting the EndpointReference at runtime).
I have also two BPEL servers installations: let´s say 1 and 2.
If I deploy these two process on the server 1, everything works fine.
If I deploy the process on the server 2, I get this error:
exception on JaxRpc invoke: HTTP transport error: javax.xml.soap.SOAPException: java.security.PrivilegedActionException: javax.xml.soap.SOAPException: Message transmission failure, response code: 500
We are trying to find out what is the diference between server 1 and server 2, as we believe they should be igual (both are 10.1.3.3.0).
And now I believe I have just found another BPEL Bug, because the log configuration on BPEL Admin seen to not work!
[]´s

Merge BPEL Servers

Similar Messages

Maybe you are looking for