Reports and summaries of the Oracle + BEA strategy are in, here is a list of the most prominent summaries:
Tuesday, July 01, 2008
Oracle BEA Strategy Briefing
Friday, June 13, 2008
World Wide Webber on Guerilla SOA
- Interview by Herbjörn Wilhelmsen here
- Jim with Martin Fowler on "Does My Bus Look Big in This?" aka "Enterprise Man Boobs" here
10.1.3.3.1 MLR#10
7013663 MLR#10 ON TOP OF 10.1.3.3.1
# **************************************************************************
# BUGS ADDED IN MLR#10, 7013663 :
# **************************************************************************
# BUG 5999832 - NULLPOINTEREXCEPTION IN DECISIONSERVICE DURING STRESS TEST OF BPEL PROCESS
# BUG 6861498 - LDAP SEARCH IN ACTIVEDIRECTORY FOR SUBGROUPS DOES NOT WORK
# BUG 6746867 - BPEL IDENTTITY SERVICE DOES NOT SOPPORT OID STRUCTURE CN <> UID
# BUG 6930522 - CUSTOMER POC ISSUE-BPEL HANDLING OF XML CDATA SECTION
# BUG 6954825 - IMPOSSIBLE TO SEE THE FLOW IF HUMAN TASK ACTIVITY CONTAINS EMAIL
# BUG 7017660 - FTPS DOES NOT WORK IN LINUX AND SOLARIS IF CIPHER SUITES ARE NOT CONFIGURED
# BUG 7010875 - JNDI LOOKUP ISSUE AS REPORTED IN BUG:6189268 IS STILL NOT INCLUDED IN THE MLRS
# BUG 6067219 - GROUP VOTING NOT WORKING
# BUG 6314009 - ESB CANNOT INVOKE EJBS - CLASSCAST EXCEPTION
# BUG 6817058 - ESB 2 WAY RS CALLING POJO THROUGH WSIF LOSES THE RETURN VALUE
# BUG 6975458 - BPELC COMPILER ERROR FOR MISSING SERVICE DEFINITION ON REMOTE HOSTED WDSL/PL
# BUG 7018925 - CORRELATION OF AQ MESSAGES NO LONGER WORKING
Monday, May 19, 2008
BPEL 10.1.3.4 Update
Here is a link to a BPEL 10.1.3.4 update webinar documentation.
Contents:
• Improved Visibility of Engine Threading Model
• Improved Statistics Collection
• Minimize XML Coding Errors
• Rescuing Lost Instances
• Automated Recovery Agent
• Collection of Support Information
• Deployment Plan
• 10.1.3.5
Friday, May 02, 2008
Now available: SOA 11g Technology Preview version 4
SOA 11g Technology Preview version 4 is out now! (May 1, 2008)
Links:
http://www.oracle.com/technology/products/jdev/11/index.html
http://www.oracle.com/technology/products/ias/bpel/techpreview/index.html
You will need a valid Metalink customer account to download it though.
Thursday, April 17, 2008
jdev.conf
Thease are my JDeveloper "performence" settings I use today:
- tools->preferences->extensions
I disabled all the extentions I do not currently use such as ADF, Webcenter ...
I edited my jdev conf file ("C:\java\jdevstudio10133\jdev\bin\jdev.conf") and added/changed the following parameteres:
AddVMOption -Dsun.java2d.noddraw=false
AddVMOption -Dsun.java2d.ddoffscreen=false
#SetJavaVM server
AddVMOption -Xmx1024M
AddVMOption -Xms512M
AddVMOption -Dcompiler.vmargs=-mx1024m
#
# Use this to prevent Windows from paging JDev's memory
# when you minimize it.
# This option should have the same effect as
# the KeepResident plug-in with the advantage
# of being a built-in feature in Sun's JVM 5.
# @see http://java2go.blogspot.com/2007/06/jdeveloper-tips-2-fine-tuning.html
AddVMOption -Dsun.awt.keepWorkingSetOnMinimize=true
Thursday, January 24, 2008
Useful links
Some links to blogentries I looked on today:
How to dynamically call a partnerlink:
"In most basic process designs, partner links are static; they refer to a single external process selected by the developer at design time. This approach is appropriate for highly targeted or constrained systems. However, in larger systems business processes are more complex. They interact with multiple external services and define multiple partner links, and some of these partner links might not be known at design time. As a result, all potential callouts and logic for deciding which partner links to use must be built inside the business process itself—unnecessarily complicating that process. Furthermore, as additional partner links are added, the resulting process grows more and more unwieldy, as any changes to the partner links require modification of the entire business process."
http://www.oracle.com/technolog
How to call a process again over and over again for error managment, caching etc
http://technology.amis.nl/blog
How to setup java.util.logging.Handler for OC4J
http://download.oracle.com/docs/cd/B31017_01/web.1013/b28952/logging.htm
http://java2go.blogspot.com/2007/11/how-to-setup-java-logging-in-oc4j.html
Wednesday, January 16, 2008
Oracle and BEA
January 16, 2008: Oracle announces an agreement to acquire BEA Systems.
Sphere: Related ContentMonday, January 14, 2008
Oracle® SOA Suite Best Practices Guide 10g Release 3 (10.1.3.3.0)
I stumbled on the document "Oracle® SOA Suite Best Practices Guide 10g Release 3 (10.1.3.3.0)" dated December 2007 today, intresting!
Sphere: Related ContentTuesday, January 08, 2008
The Next-Generation, Grid-Enabled Service-Oriented Architecture
The following two articles gives an insight on how the Tangosol Coherence product has been utilized in the 11g version of Oracle SOA Suite ( I hope )
The Next-Generation, Grid-Enabled Service-Oriented Architecture
Next-Generation Grid-Enabled SOA: Not Your MOM's Bus
Oracle SOA Suite 11 g - Technology Preview
Here is the link to Oracle SOA Suite 11g - Technical Preview 3
Happy new 2008!
Thursday, December 13, 2007
Patch 6501312 - BPEL Repository spring cleaning
A new patch with id 6501312 creates a purge_instances procedure in the oraclebpel schema that you can use to delete your old BPEL Process instances.
Currently (Thu Dec 13 15:36:56 CET 2007) there seemes to be a problem with the paths for this patch in unix so I installed it using the following procedure:
1 Copy the files in 6501312/files/bpel\system\database\scripts to $ORACLE_HOME/bpel/system/database/scripts
2. Run the sqlplus scrip to create index and create the procedure "purge_instances".
Example:
bash$ sqlplus /nolog
SQL*Plus: Release 10.1.0.5.0 - Production on Thu Dec 13 10:19:06 2007
Copyright (c) 1982, 2005, Oracle. All rights reserved.
SQL> connect orabpel@instance
Enter password:
Connected.
SQL> @upgrade_6501312_oracle.sql
SQL> commit;
SQL> @purge_instances_oracle.sql
After this you may start your spring cleaning:
BEGIN
purge_instances(to_timestamp('2007-12-13 10:30:25','YYYY-MM-DD HH24:MI:SS'));
COMMIT;
END;
Happy cleaning!
Sphere: Related Content[Error ORABPEL-10902]: compilation failed, XML parsing failed because ""
ORABPEL-10902 shows up from time to time to annoy any BPEL developer.
It could be (I think) somthing like:
- General namespace errors
- Incorrect variable errors
- Variables that you try to access "out-of-scope"
The error message "" is not that helpful in any case, but if we change the 'verbose' setting to true in the ant build.properties file we get better information.
deployment tasksverbose = true
Thanks to Marc Kelderman, I can't wait until I get the next ORABPEL-10902!
Today we saw an ORA-600 when we added a diskgroup into our ASM instance... but that is a different story.
Sphere: Related ContentWednesday, December 12, 2007
Trick Question - "good but expensive" vs "quick and dirty"
Your project manager has asked you to estimate the difference in cost between the "good but expensive", "quick and dirty" path or solution!
What should you do?
Steve McConnells answer in this blogentry is that you should ask three more questions:
1.How much it will cost to backfill the good path after you've already gone down the quick path?
2.If you choose the quick path now, how much does that slow down other work until you're able to retrofit the good path?
3.Is there a path that is quicker than the good path and that won't affect the rest of the system?
Steves Bottom line: When facing the prospect of taking on technical debt, be sure to generate more than two design options.
10.1.3.3.1 deployment
After deploying 10.1.3.3.1 I got the following startup errors (even if I followed the README to the point)
--------
07/12/11 22:10:41 Start process--------
07/12/11 22:11:39 log4j:WARN No appenders could be found for logger (wsif).
07/12/11 22:11:39 log4j:WARN Please initialize the log4j system properly.¨
07/12/11 22:11:44 WARNING: Application.setConfig Application: orabpel is in failed state as initialization failed. java.lang.InstantiationException: Error initializing ejb-modules: error in opening zip file; nested exception is: java.util.zip.ZipException: error in opening zip file
07/12/11 22:11:44 WARNING: Application.setConfig Application: hw_services is in failed state as initialization failed. java.lang.InstantiationException: Error initializing ejb-modules: [hw_services:hw_services_ejb:RemoteWorklistService] - Unable to load ejb-class oracle.tip.pc.services.hw.worklist.RemoteWorklistServiceBean, see section 23.2 of the EJB 2.1 specificationoracle.classloader.util.AnnotatedClassNotFoundException:
Missing class: oracle.tip.pc.services.hw.worklist.RemoteWorklistServiceBean
Dependent class: com.evermind.server.ejb.deployment.BeanDescriptor Loader: oc4j:10.1.3 Code-Source: /oracle/product/soa_10.1.3/j2ee/home/lib/oc4j-internal.jar Configuration:
This load was initiated at hw_services.root:0.0.0 using the Class.forName() method.
The missing class is available from the following locations:
1. Code-Source: /oracle/product/soa_10.1.3/bpel/system/services/lib/bpm-services.jar (from
07/12/11 22:11:50 Initializing B2BDiscoveryServlet07/12/11 22:11:54 Finished initializing B2BDiscoveryServlet07/12/11 22:11:56 Oracle Containers for J2EE 10g (10.1.3.3.0) initialized
I then tried to redeploy BPEL services by using the $ORACLE_HOME/config/deploy_bpel.sh script.
This script gave me the following error:
config>> ./deploy_bpel.sh
BUILD FAILED/oracle/product/soa_10.1.3/bpel/system/services/install/ant-tasks/redeploy.xml:33: The following error occurred while executing this line:/oracle/product/soa_10.1.3/bpel/system/services/install/ant-tasks/redeploy.xml:66: Error while expanding /oracle/product/soa_10.1.3/bpel/system/services/lib/workflow_javadoc.jar
I then commeted out the workflow_javadoc.jar from bpel/system/services/install/ant-tasks/redeploy.xml, and then the redeploy works again!
Tuesday, December 11, 2007
10.1.3.3.1 Patchset Release now available!
Patchset 10.1.3.3.1 for 10.1.3.3.0 is avaliable as Patchid 6492514.
For more information see the BPEL home page.
Monday, November 26, 2007
Teamwork: Feedback
In software development teams it is important to be open and give feedback. I have been trained to give feedback with my feelings, instead of my thoughts, and to focus on a behaviour in a "I" perspective (Not "the group thinks ").
The structure of the feedback is:
When you ...
I feel that ...
This causes ...
I wish that ...
The "I Feel" makes it hard to argue is a true statement...
I also think that it is important to educate my team on feedback reactions.
The Feedback reaction staircase looks likes follows (I guess i neeed an image for this )
___________________
___________|Continue or Change
________|Understand
_______|Explain
_______|Defend
Reject|
The idea here is that it is important to be able to "catch" feedback in the two uppermost steps, ie it is better to Understand the feedback than to Reject, Defend or Explain your behaviour. It is very,very easy to start explaining your behaviour.
Examples:
Change – ”I think that was good feedback, I will try to change my behaviour, could you help?"
Continiue – ”I think I have undersood your feedback but I will continue as before because ... "
Understand – ”Yes I understand. Could you explain a bit further.."
Explain – ”Yes, but it was because I was tired”
Defend – ”No, you misunderstood”
Reject – ”I Do not care!"
(This is from the ugl.biz courses)
Thursday, November 22, 2007
Purging runtime documents in B2B
The Oracle B2B purge utilities are used as :
%. setenv.sh (b2b env)
% java -ms1024M -mx1024M oracle.tip.repos.purge.PurgeManager purgeRuntime -start 01-JAN-2007 -end 20-AUG-2007 STATUS
Execution context started.
Purge started, please wait...
Purge completed.
Or with PL/SQL
exec purge_runtime('01-JAN-2007',20-AUG-2007','STATUS
Where STATUS
B2B and Tracking Messages in AQ
This thread on Tracking Messages in AQ in the Oracle forums is intresting
Ramesh Nittur Anantharamaiah suggests the following to for a status overview:
SELECT a.queue_table, a.name, b.waiting, b.ready, b.expired
FROM dba_queues a, v$aq b
WHERE a.qid = b.qidAND owner = 'B2B'
Martien van den Akker suggests the following query to be able to see the message ids easily:
SELECT qtb.queue,
qtb.msg_id,
qtb.msg_state,
qtb.enq_time,
qtb.delay,
qtb.user_data.msg_id,
qtb.user_data.inreplyto_msg_id b2b_inreplyto_msg_id,
decode(instr(qtb.user_data.inreplyto_msg_id, ':'), 0, qtb.user_data.inreplyto_msg_id, SUBSTR(qtb.user_data.inreplyto_msg_id, 1, instr(qtb.user_data.inreplyto_msg_id, ':') -1)) inreplyto_msg_id,
decode(instr(qtb.user_data.inreplyto_msg_id, ':'), 0, NULL, SUBSTR(qtb.user_data.inreplyto_msg_id, instr(qtb.user_data.inreplyto_msg_id, ':') + 1)) conversation_id,
qtb.user_data.from_party,
qtb.user_data.to_party,
qtb.user_data.action_name,
qtb.user_data.doctype_name,
qtb.user_data.doctype_revision,
qtb.user_data.msg_type,
qtb.user_data.payload,
qtb.user_data.attachment,
qtb.consumer_namefrom aq$ip_qtab qtb
ORDER BY enq_time DESC;
I also noted that if you have GridControl 10.2 you can create a UDM based on the following query (reference to IOUG Collaborate 07: Oracle 10g Grid Control ) :
select q.owner '.'q.name queue_name, s.readyfrom v$aq s, dba_queues q where s.qid = q.qidand q.owner IN ('B2B')
And then set alert levels as you wish:
Warning "B2B.IP_OUT_QUEUE:50;B2B.IP_IN_QUEUE:50"
Error "B2B.IP_OUT_QUEUE:100;B2B.IP_IN_QUEUE:100"
And then create create error messages such as "Queue %key% contains %value% unattended messages."
Today I also use this query to have a peek in the actual messages and "DUMP" the contents to be able to try to understand a charachter set issue:
SELECT qtb.queue,
qtb.msg_id,
DBMS_LOB.SUBSTR(qtb.user_data.payload,4000,1),
DUMP (DBMS_LOB.SUBSTR(qtb.user_data.payload,4000,1),16)
FROM aq$ip_qtab qtb
WHERE MSG_ID = '3989F49C24206660E0440003BAE85389';
To list the consumers registrated for the B2B in and out queues one may use the following SQL:
SELECT *
FROM aq$ip_qtab_s
Subscribe to:
Posts (Atom)
