BEA WebLogic Integration Release Notes

Change Request Number	Description
CR173851	Deadlocks may sometimes occur during event processing. The event processor has built in retry capabilities to take care of this situation. Now, the number of retires and the delay between retries may be specified on the command line as bpm.eventlistener.MaxDeadlockRetries and bpm.eventlistener.DeadlockDelay JVM environment variables. The deadlock delay is the number of milliseconds the thread will sleep before attempting to reprocess the event. Example: -Dbpm.eventlistener.MaxDeadlockRetries=5 -Dbpm.eventlistener.DeadlockDelay=750 In this example, the thread will sleep for 750 milliseconds after encountering a deadlock and then attempt to process the event again. This will be repeated 5 times, for a total of 6 attempts (the original plus the 5 retries). The default values for these variables are 3 and 1000 respectively.
CR189762	When there is a database failure, some ebXML public workflows may not start even after the database recovers. When this problem occurs, the client trading partner appears to be "stuck" while attempting the initial B2B conversation. In addition, the client trading partner may also receive a 400 HTTP status. In such a scenario, ensure that the client trading partner posts a new request and performs whatever cleanup that is necessary to abandon the old request.

Change Request
Number

Description

CR173851

Deadlocks may sometimes occur during event processing. The event processor has built in retry capabilities to take care of this situation.

Now, the number of retires and the delay between retries may be specified on the command line as bpm.eventlistener.MaxDeadlockRetries and bpm.eventlistener.DeadlockDelay JVM environment variables.

The deadlock delay is the number of milliseconds the thread will sleep before attempting to reprocess the event. Example:

-Dbpm.eventlistener.MaxDeadlockRetries=5

-Dbpm.eventlistener.DeadlockDelay=750

In this example, the thread will sleep for 750 milliseconds after encountering a deadlock and then attempt to process the event again. This will be repeated 5 times, for a total of 6 attempts (the original plus the 5 retries).

The default values for these variables are 3 and 1000 respectively.

CR189762

When there is a database failure, some ebXML public workflows may not start even after the database recovers. When this problem occurs, the client trading partner appears to be "stuck" while attempting the initial B2B conversation. In addition, the client trading partner may also receive a 400 HTTP status.

In such a scenario, ensure that the client trading partner posts a new request and performs whatever cleanup that is necessary to abandon the old request.

The database table: EVENTKEY does not contain the columns: VALUEMATCHONLY. Please consult your database mappings in the weblogic-cmp-rdbms.xml deployment descriptor and ensure these match your database schema.

Users of this feature . . .	Can migrate to . . .
XOCP protocol	ebXML or RosettaNet
XOCP BPM plug-in
cXML protocol
B2B logic plug-ins	No equivalent functionality
Messaging API	ebXML JMS interface or BPM workflows
Trading partner zeroweight client	A JSP tag library with Portlet and HTML-based Zeroweight trading partner clients will be made available in a future release.
Swing-based BPM Worklist client	A JSP-based Worklist client is available with the WebLogic Integration SP6 release. Note: Only the current Java Swing user interface is being deprecated. The Worklist API is not being deprecated.

Users of this feature . . .

Can migrate to . . .

XOCP protocol

ebXML or RosettaNet

XOCP BPM plug-in

cXML protocol

B2B logic plug-ins

No equivalent functionality

Messaging API

ebXML JMS interface or BPM workflows

Trading partner zeroweight client

A JSP tag library with Portlet and HTML-based Zeroweight trading partner clients will be made available in a future release.

Swing-based BPM Worklist client

A JSP-based Worklist client is available with the WebLogic Integration SP6 release.

Note: Only the current Java Swing user interface is being deprecated. The Worklist API is not being deprecated.

Change Request Number	Description
CR121357	TimeListenerBean onMessage does not remove all WorkflowProcessor beans. The problem is that the workflow processor bean instance is not removed inside this do..while loop, and the workflowprocessor.remove() call is called only once when exiting the onMessage method. This means that, in a situation where instanceid does not have a matching workflow instance, 20 WorkflowProcessor bean instances are created and only the last one of them is removed. This problem was resolved with an engineering code change.
CR122714	RDBMS adapter throws exception outside of the <eda> structure when: a) the incorrect user name/password is entered b) the incorrect URL is entered c) a non-existing stored procedure is used Customer needs to see the adapter return an <eda> XML structure under all circumstances. Any error code and information should be returned within this structure. This problem was resolved with an engineering code change.
CR124683 CR182704	Customer has two cluster domains: domain1 and domain2. Each domain has two managed servers; each managed server has two JMS servers (one for WLI-specific work and another for the JMS messaging bridge). The sending application sets the JMSReplyTo field as a distributed destination on domain1, and a message travels over the messaging bridge and ends up in the JMSStore on domain2. In the meantime, the domain2 managed server is killed by Kill-9. When the managed server restarts, the JMSServer attempts to locate the ConfigMbeanName for the distributed destination member instance that was written to the JMSStore in domain2 but cannot locate this name because it belongs to domain1 and the JMSServer is throwing a NullPointerException. This problem was resolved with an engineering code change.
CR129512	The customer has requested additional logging functionality to include: a) a log of the input XML message from the event processor for a start or event node b) a log of a timer message when it fires The added functionality was provided with an engineering code change.
CR133445	The following intermittent license error occurs at run time: java.lang.RuntimeException: License class files have been tampered. This problem was resolved with an engineering code change.
CR134388	The Workflow label is not getting updated in the worklist. This problem was resolved with an engineering code change.
CR135860	Customer has two configured queues: com.bea.wlpi.EventQueue to send mex to WLI com.bea.wlpi.EventQueueExt to receive mex from WLI To send the correct answer to the client, the customer is using the attribute JMSCorrelationID in the JMS mex. The activated workflow takes the attribute JMSCorrelationID from the returned mex to use in the answer. This does not work and the client continues to wait. This problem was resolved with an engineering code change.
CR181920	TimeListenerBean consumes a failed message when the WLI database is not running. As of 7.0 sp6, a new system property allows timed events to be rolled back for JMS re-delivery, allowing timed events to be recovered in the event of a database crash. To turn on the feature, set: -Dwli.bpm.server.timeprocessor.rollbackafterretry=true Note however that if workflows are designed such that it is possible for there to be timers pending when the workflow completes, and the completed instance is deleted before the timer message is delivered, the timer messages will be retried by JMS. These event messages would end up in the failure queue when, in fact, they should have been consumed.
CR182556	CDATA is not properly handled or parsed by the XPATH expression inside the workflow. The problem was fixed with an engineering code change whereby the CDATA value appears properly when being handled by the XPATH expression in a workflow.
CR182709	While analyzing system performance, a lot of threads were waiting in synchronization in XMLParserHelper class. This problem was resolved with an engineering code change.
CR183277	Timed event is performed abnormally when it is reactivated or edited. Timed started workflows are triggered whenever that workflow is saved in studio. Time started workflows should start only when their "Start Date Expression" specifies they should start. This problem was resolved with an engineering code change.
CR184894	Over-synchronization occurs when moving messages from the AI to BPM queue. This problem was resolved with an engineering code change.
CR187045	Sometimes when Oracle listener is killed and re-started later, the following exception is seen: Setting Transaction isolation to TRANSACTION_READ_COMMITTED. This problem was resolved with an engineering code change.
CR188141	Issues occurred when adding re-routes using Java API. The Java API that is being used is similar to <WLI_HOME>/samples/bpm_api/commandline/CLStudio.java. Engineering provided a new debug flag for re-routes. It is controlled with the following boolean Java system property: com.bea.wlpi.reroute.debug
CR188356	TimedEvent failed to continue after Oracle server was rebooted. This problem was resolved with an engineering code change.
CR193316	A receiver stops processing messages when a RedeliveryLimit is configured and the number of times the RedeliveryLimit is reached is greater than the MessagesMaximum setting on the ConnectionFactory. For example: RedeliveryLimit=0, MessagesMaximum=10. Receiver receives a message and calls session.recover() 10 times. The receiver will stop processing messages and the console will show 10 pending messages. Code was added to adjust the window counter when a message is removed from the queue because the RedeliveryLimit had been reached.
CR193661	iWay Adapter (FILE & RDBMS) Application views Deployment/Undeployment is sometimes unreliable. This problem was resolved with an engineering code change.
CR200912	The BPMAdmin utility was unable to export a business calendar and was giving the following exception: javax.ejb.CreateException: The system could not find the specified business calendar. This problem was resolved with an engineering code change.
CR202956	The WLI HTTP plugin randomly throws the following exception: no templatedef found. The exception occurs when the post events pumped are concurrent and high in number to a workflow, which is configured to start on a HTTP event. This prevents the workflow from processing further events. This problem was fixed with an engineering code change containing a replacement WAR and JAR for the HTTP plugin.

Change Request
Number

Description

CR121357

TimeListenerBean onMessage does not remove all WorkflowProcessor beans. The problem is that the workflow processor bean instance is not removed inside this do..while loop, and the workflowprocessor.remove() call is called only once when exiting the onMessage method. This means that, in a situation where instanceid does not have a matching workflow instance, 20 WorkflowProcessor bean instances are created and only the last one of them is removed. This problem was resolved with an engineering code change.

CR122714

RDBMS adapter throws exception outside of the <eda> structure when:

a) the incorrect user name/password is entered

b) the incorrect URL is entered

c) a non-existing stored procedure is used

Customer needs to see the adapter return an <eda> XML structure under all circumstances. Any error code and information should be returned within this structure. This problem was resolved with an engineering code change.

CR124683

CR182704

Customer has two cluster domains: domain1 and domain2. Each domain has two managed servers; each managed server has two JMS servers (one for WLI-specific work and another for the JMS messaging bridge). The sending application sets the JMSReplyTo field as a distributed destination on domain1, and a message travels over the messaging bridge and ends up in the JMSStore on domain2. In the meantime, the domain2 managed server is killed by Kill-9. When the managed server restarts, the JMSServer attempts to locate the ConfigMbeanName for the distributed destination member instance that was written to the JMSStore in domain2 but cannot locate this name because it belongs to domain1 and the JMSServer is throwing a NullPointerException. This problem was resolved with an engineering code change.

CR129512

The customer has requested additional logging functionality to include:

a) a log of the input XML message from the event processor for a start or event node

b) a log of a timer message when it fires

The added functionality was provided with an engineering code change.

CR133445

The following intermittent license error occurs at run time: java.lang.RuntimeException: License class files have been tampered. This problem was resolved with an engineering code change.

CR134388

The Workflow label is not getting updated in the worklist. This problem was resolved with an engineering code change.

CR135860

Customer has two configured queues:

com.bea.wlpi.EventQueue to send mex to WLI

com.bea.wlpi.EventQueueExt to receive mex from WLI

To send the correct answer to the client, the customer is using the attribute JMSCorrelationID in the JMS mex. The activated workflow takes the attribute JMSCorrelationID from the returned mex to use in the answer. This does not work and the client continues to wait. This problem was resolved with an engineering code change.

CR181920

TimeListenerBean consumes a failed message when the WLI database is not running. As of 7.0 sp6, a new system property allows timed events to be rolled back for JMS re-delivery, allowing timed events to be recovered in the event of a database crash. To turn on the feature, set: -Dwli.bpm.server.timeprocessor.rollbackafterretry=true

Note however that if workflows are designed such that it is possible for there to be timers pending when the workflow completes, and the completed instance is deleted before the timer message is delivered, the timer messages will be retried by JMS. These event messages would end up in the failure queue when, in fact, they should have been consumed.

CR182556

CDATA is not properly handled or parsed by the XPATH expression inside the workflow. The problem was fixed with an engineering code change whereby the CDATA value appears properly when being handled by the XPATH expression in a workflow.

CR182709

While analyzing system performance, a lot of threads were waiting in synchronization in XMLParserHelper class. This problem was resolved with an engineering code change.

CR183277

Timed event is performed abnormally when it is reactivated or edited. Timed started workflows are triggered whenever that workflow is saved in studio. Time started workflows should start only when their "Start Date Expression" specifies they should start. This problem was resolved with an engineering code change.

CR184894

Over-synchronization occurs when moving messages from the AI to BPM queue. This problem was resolved with an engineering code change.

CR187045

Sometimes when Oracle listener is killed and re-started later, the following exception is seen: Setting Transaction isolation to TRANSACTION_READ_COMMITTED. This problem was resolved with an engineering code change.

CR188141

Issues occurred when adding re-routes using Java API. The Java API that is being used is similar to <WLI_HOME>/samples/bpm_api/commandline/CLStudio.java.

Engineering provided a new debug flag for re-routes. It is controlled with the following boolean Java system property: com.bea.wlpi.reroute.debug

CR188356

TimedEvent failed to continue after Oracle server was rebooted. This problem was resolved with an engineering code change.

CR193316

A receiver stops processing messages when a RedeliveryLimit is configured and the number of times the RedeliveryLimit is reached is greater than the MessagesMaximum setting on the ConnectionFactory.

For example: RedeliveryLimit=0, MessagesMaximum=10. Receiver receives a message and calls session.recover() 10 times. The receiver will stop processing messages and the console will show 10 pending messages.

Code was added to adjust the window counter when a message is removed from the queue because the RedeliveryLimit had been reached.

CR193661

iWay Adapter (FILE & RDBMS) Application views Deployment/Undeployment is sometimes unreliable. This problem was resolved with an engineering code change.

CR200912

The BPMAdmin utility was unable to export a business calendar and was giving the following exception:

javax.ejb.CreateException: The system could not find the specified business calendar.

This problem was resolved with an engineering code change.

CR202956

The WLI HTTP plugin randomly throws the following exception: no templatedef found. The exception occurs when the post events pumped are concurrent and high in number to a workflow, which is configured to start on a HTTP event. This prevents the workflow from processing further events. This problem was fixed with an engineering code change containing a replacement WAR and JAR for the HTTP plugin.

1	CR063709 CR075768
	Problem	The RosettaNet Security Sample throws an exception when a single database is used for two peers and the database tables for peer1 are overwritten by peer2. The default database for the RosettaNet 2.0 Security sample cannot be changed easily from PointBase. Unlike the other B2B samples, the RosettaNet 2.0 Security sample does not support the use of the WebLogic Integration Database Wizard to switch the database to be used with the samples domain.
	Platforms	All
	Workaround	If possible, use different databases for different peers. For example, have peer1 use MSSQL and peer2, Oracle. If you must run two peers on one machine and both peers use the same database server, make sure that: With MSSQL—A different database name is used for each peer, such as WLIDB for peer1 and WLIDB2 for peer2. With Oracle—A different Oracle user account is used for each peer, such as kevin/kevin for peer1 and nina/nina for peer2. Specify every database you use for this sample on the RunRN2Security command line, as follows: RunRN2Security database_for_peer1 database_for_peer2 Edit the setDBVars and setDBVarsExt files to set or change the database connection parameters. The setDBVars and setDBVarsExt files are located in the following locations: SAMPLES_HOME/integration/config/samples/ RN2Security/config/peer1/dbInfo/database_type SAMPLES_HOME/integration/config/samples/ RN2Security/config/peer2/dbInfo/database_type In both pathnames SAMPLES_HOME is the samples installation directory and database_type is the type of database to which you are connecting. Database types are oracle, mssql, pointbase, and sybase.
2	CR076994
	Problem	In a cluster, both the Administration and Managed Servers require installation of WebLogic Integration.
	Platforms	All
	Workaround	WebLogic Integration users must install WebLogic Integration on the Administration Server as well as the Managed Servers.
3	CR079611
	Problem	In the Studio Routing window, when creating a new WebLogic Server group and mapping a new BPM role to that group, the list of roles does not update with the new role.
	Platform	Windows
	Workaround	Shutdown Studio and restart it; the new role will be visible.
4	CR081119
	Problem	The default installation includes the protocol definition for RN-Hub. However, RN-Hub support for RosettaNet is not supported.
	Platforms	All
	Workaround	None
5	CR090195
	Problem	The Event list in the Start and Event node is not refreshed when a plug-in is loaded or unloaded while workflows that do not refer the plug-in is open.
	Platforms	Windows
	Workaround	Close all open Workflow templates before making plug-in state change.
6	CR092283
	Problem	Error when creating a BPM Plug-In for AI in an existing domain. The configDomain script fails when the BPM Plug-In already exists in the domain's config.xml file: <WLI_HOME>\project\ai\ai_resources.xml:295: Unresolved module reference encountered: EJBComponent:WLI-BPM Server in ModuleComponent/BPM_Plugin This error occurs because the configDomain script adds an entry of BPM Plug-in for Application Integration in config.xml.
	Platforms	Windows
	Workaround	Do not to select the BPM Plug-in if it already exists in a config.xml file of a domain in which it is being created.
7	CR093191
	Problem	WebLogic Workshop users cannot create an application view control for the 7.0 SP2 RDBMS adapter.
	Platform	All
	Workaround	Please contact BEA Customer Support.
8	CR094433
	Problem	Incomplete workflow instances may occur if JTA and JMS migration is done using HTTP protocol for the failover.
	Platform	All
	Workaround	Use the t3 protocol.
9	CR094470
	Problem	If config.xml contains <StartupClass> or <ShutdownClass> modules, the configDomain script adds a LoadOrder attribute to it, which will make the server unbootable.
	Platform	All
	Workaround	In config.xml, remove the LoadOrder attribute entry for the <ShutdownClass> and <StartupClass> before starting the server: <ShutdownClass Arguments="mode=TERMINATE" LoadOrder="800" ClassName="com.bea.b2b.server.Shutdown" Name="WLCShutdown" Targets="myserver"/> <StartupClass ClassName="com.bea.lwclient.Startup" Name="LwcStartup" Targets="myserver"/>
10	CR095276
	Problem	In Domain Wizard generated EAI, BPM, WLI domains, the default web application directory name uses myserver when users give their own name for an admin server. This happens during the domain generation and the admin server fails to boot with following error: <Error> <Management> <141006> <Application eaiC:Name=DefaultWebApp_<user_specified_name>, Type=Application not found at <user_domain>\applications\ DefaultWebApp__<user_specified_name>>
	Platforms	All
	Workaround	Change the directory name from applications/DefaultWebApp_myserver to DefaultWebApp_<user_specified_name> under applications/. In config.xml, change any DefaultWebApp_myserver to DefaultWebApp_<user_specified_name>.
11	CR095278
	Problem	Existing workflows using Application Integration do not work in the sample DBMS Adapter.
	Platforms	All
	Workaround	For complex applications, re-export the whole set of related objects that go with the workflow and re-import. For simple applications, change the request XML documents to uppercase.
12	CR096379
	Problem	If you are using RDBMS Realm, the SchemaProperties attribute of RDBMSRealm element in the config.xml of the Domain Wizard generated wlidomain is incomplete.
	Platforms	All
	Workaround	Append the following text to the end of the line starting with "SchemaProperties=", right after "PERMISSION FROM AC", without a space. PERMISSION FROM ACLENTRIES WHERE PERMISSION = ?; getUser=SELECT USERID, PASSWORD FROM WLSUSER WHERE USERID = ?;deleteGroup2=DELETE FROM ACLENTRIES WHERE PRINCIPAL = ?;deleteGroup1=DELETE FROM GROUPMEMBER WHERE GROUPID = ?;deleteUser1=DELETE FROM ACLENTRIES WHERE PRINCIPAL = ?;getAclEntries=SELECT NAME, PRINCIPAL, PERMISSION FROM ACLENTRIES WHERE NAME = ? ORDER BY PRINCIPAL;getGroupMembersGroups=SELECT GROUPMEMBERID, GROUPID FROM GROUPMEMBER WHERE GROUPID = ?;getGroups=SELECT GROUPID FROM WLSGROUP ORDER BY GROUPID;getGroup=SELECT GROUPID FROM WLSGROUP WHERE GROUPID = ?;getUsers=SELECT USERID, PASSWORD FROM WLSUSER ORDER BY USERID"/>
13	CR119473
	Problem	In the Sample Applications menu page of the samples launcher, selecting the link to the BPM-Workshop sample causes a browser error (Error 404--Not Found) on UNIX systems due to case sensitivity issues. The Sample Applications menu page is located as follows: SAMPLES_HOME/integration/config/samples/applications/ DefaultWebApp_myserver/menu.html
	Platforms	UNIX
	Workaround	Rename the BpmWlwSampleDesc.html file to BPMWLWSampleDesc.html in the following directory: SAMPLES_HOME/integration/config/samples/applications/ DefaultWebApp_myserver

<weblogic-enterprise-bean>
<ejb-name>WorkflowInstance</ejb-name>
<entity-descriptor>
<entity-cache>
<max-beans-in-cache>100</max-beans-in-cache>
<idle-timeout-seconds>600</idle-timeout-seconds>
<concurrency-strategy>Database</concurrency-strategy>
<cache-between-transactions>False</cache-between-transactions>
</entity-cache>

<transaction-isolation>
<isolation-level>TRANSACTION_READ_COMMITTED_FOR_UPDATE</isolation-level>
<method>
<ejb-name>WorkflowInstance</ejb-name>
<method-name>*</method-name>
</method>
</transaction-isolation>

$WLI_HOME/setenv.sh
CLASSPATH=
$WL_HOME/lib/weblogic.jar:$WLI_HOME/lib/wlpi-ejb.jar:
$WLI_HOME/lib/sampleplugin-ejb.jar
$JAVA_HOME/bin/java -classpath "$CLASSPATH"
com.bea.wlpi.tour.po.plugin.StartOrderDriver
t3://localhost:7001 joe password

Table 4 provides the missing information. For example, the first row in the table indicates that template definitions created in WebLogic Integration 2.1 SP2 are usable in all versions, whereas the second row indicates that a template definition created in WebLogic Integration 7.0 GA is not usable in the WebLogic Integration 2.1 SP2 release.

Version	Compatible with
	2.1 SP2	7.0 GA	7.0 SP1	7.0 SP2	7.0 SP4	7.0 SP5	7.0 SP6
2.1 SP2	Yes	Yes	Yes	Yes	Yes	Yes	Yes
7.0 GA	No	Yes	Yes	Yes	Yes	Yes	Yes
7.0 SP1	No	Yes	Yes	Yes	Yes	Yes	Yes
7.0 SP2	No	No*	No*	Yes	Yes	Yes	Yes
7.0 SP4	No	No*	No*	Yes	Yes	Yes	Yes
7.0 SP5	No	No*	No*	Yes	Yes	Yes	Yes
* This is due to a bug in 7.0 GA and 7.0 SP1.

1	CR196373
	Problem	WebLogic Integration B2B Server may sometimes fail to start on managed server recovery. The following entries are then made in the log file: <Timestamp> <Error> <B2B> <000000> <<B2BIntegration-Server> ERROR: WebLogic Integration - B2B Server failed to start with null.> <Timestamp> <Error> <B2B> <000000> <<B2B Server> ERROR: WebLogic Integration - B2B Server failed to start with com.bea.b2b.server.InitException.> Usually, this failure is preceded by some OracleXAResource.checkError exceptions. However, you may ignore these exceptions. The server will continue to start; however, no new B2B messages will be accepted. Recovered ebXML messages will fail with the following error: <Timestamp> <Error> <B2B> <ceraphin.bea.com> <managed2_hwbwl98> <ExecuteThread: '12' for queue: 'default'> <kernel identity> <> <000000> <<EBXML-Protocol> ERROR: Exception During send message bcManageProvideOrder-86420-1097604825564-616-86420-1097605014693-408: java.lang.NullPointerException.>
	Platforms	All
	Workaround	Restart the server within the B2B and JMS retry window so that full recovery can take place.
2	CR193997
	Problem	It can take several minutes for in-doubt transactions to show up in Oracle. There may be a race if recovery has started prior to Oracle detecting the loss of the transaction messages.
	Platforms	All
	Workaround	Wait several minutes before starting recovery either by restarting the server or by doing a JTA migration. After initiating recovery, if you check the Oracle dba_2pc_pending table and see a record associated with the failed server that has a timestamp earlier than the time when recovery was initiated, recovery will fail. You must then restart the server.

14	CR134353
	Problem	When a WebLogic Integration domain is XA-enabled an oracle.jdbc.xa.OracleXAException may be thrown from oracle.jdbc.xa.client.OracleXAResource.commit calls during a server restart after a crash. The exception is benign and indicates the attempted recovery of transactions that committed before the crash happened but before the server transaction logs got a chance to record these commits.
	Platforms	All
	Workaround	None
15	CR135996
	Problem	In the BPM-Workshop inter operability sample, setting the ListenAddress to localhost causes the test to fail and show incomplete links.
	Platforms	All
	Workaround	Leave the ListenAddress blank when configuring the platform domain.
16	CR137004
	Problem	The XML response from events and services may contain irrelevant values when the isNull Attribute is retrieved as true.
	Platforms	All
	Workaround	None, ignore these values.
17	CR177271
	Problem	During the server startup, informational messages about SAXParserExceptions from malformed web.xml for the HTTP plugin may be seen in the server logs.
	Platforms	All
	Workaround	None, ignore these values. They are benign. These exceptions have no effect on the functionality of the HTTP plug-in.
18	CR125694
	Problem	RunSamples does not run on Windows XP.
	Platforms	Windows XP
	Workaround	Add -Dos.name="Windows 2000" to the java start up line in the script.