Home
 

User login

 
 

Navigation

 
 

Events

« May 2008
SunMonTueWedThuFriSat
123
45678910
11121314151617
18192021222324
25262728293031
 

Cape Clear Server 7.0.1 Release Notes

Cape Clear Server 7.0.1 Release Notes are divided into two sections:

Known Issues in Cape Clear Server 7.0.1

Issue Summary: Database issue: FKCBF1E30FE39796AF constraint on ACTIVITY table
ID: SVR-2432
Component: Runtime -BPEL Runtime
Description:

The database creation scripts included with Cape Clear Server add a foreign key constraint on the ACTIVITY table called FKCBF1E30FE39796AF, referencing the THREAD_CURSOR table.
This constraint is occasionally violated in some recovery scenarios. It is not necessary and can be safely removed if it causes issues. The following SQL statement will accomplish this:

alter table ACTIVITY drop constraint FKCBF1E30FE39796AF;

You may encounter this issue if:

  • You have installed 7.0.0 or 7.0.1 and continue to use the database that you created for those releases, even if you have upgraded Cape Clear Server to a later patch release.
  • If you migrated 6.7 databases using the 7.0.0 or 7.0.1 migration script.

 

Issue Summary: Increased PermGen Size Required for WebLogic
ID: SVR-2437
Component: Integrations - WebLogic
Description: When running Cape Clear Server deployed within WebLogic, a "java.lang.OutOfMemoryError: PermGen space" failure may occur. To ensure that the JVM permanent generation space will not be exhausted, configure a JVM startup parameter of -XX:MaxPermSize=512m in the WebLogic
$CAPECLEAR_DOMAIN_HOME/bin/setDomainEnv.cmd file (for Windows) or $CAPECLEAR_DOMAIN_HOME/bin/setDomainEnv.sh (for UNIX platforms).

 

Issue Summary: "java.lang.OutOfMemoryError: PermGen space failure" when editing Orchestration manager web.xml
ID: SVR-2171
Component: Administration - Orchestration Manager
Description: Editing the Orchestration Manager web.xml file for a running server will lead to a redeployment of the Orchestration manager web application. That redeployment may lead to a "java.lang.OutOfMemoryError: PermGen space failure"

The most common cause for that error is that you simply don't have enough space to load up all your classes. Use the param '-XX:MaxPermSize=128m' to increase the memory available.

Issue Summary: Routes in a comma separated list of routes in a clustered route cannot have spaces after commas
ID: SVR-2194
Component: Administration - Manager
Description: When adding a route to the name.of.routes setting, you cannot use comma after the names. For example, if you created two HTTP routes named testroute1, testroute2 and then tried to create a cluster route whose setting for name.of.routes is"testroute1, testroute2" fails while "testroute1,testroute2" works. You need to strip whitespace after the comma.


Issue Summary: The Adapters Template page in Cape Clear Manager displays JMS adapter templates, which are Retired Features
ID: SVR-2239
Component: Administration - Manager
Description: The Adapters Template page in Cape Clear Manager displays JMS adapter templates, which are Retired Features. The following items should be removed from the Cape Clear Manager JSPs:

* JMS JBossJMS@localhost JMS provider JBoss 4.0 JBoss 4.0 channel configuration
* JMS WebLogicJMS@localhost WebLogic JMS Provider 9.1 WebLogic JMS example configuration for localhost
* JMS WebSphereMQ@localhost


Issue Summary: BPEL timer events should deal in UTC (Zulu time) to avoid daylight savings switch over issues
ID: SVR-2269
Component: Documentation - Release Notes - Issues Outstanding
Description: There is a bug in the way we handle duration based alarms and waits that start before the switch over time for daylight savings and that expire afterwards. Currently these calculations are done using local time and disregard the time changeover.
Deadline based alarms and waits are not affected by this issue.

Example:
Imagine that clocks go forward 1 hour at 1:00am. If we start a wait of duration 1 hour at 0:30am, we will calculate that the wait should expire at 1:30am. At 01:00, the system clock goes forward 1 hour and the wait is now treated as expired. That is, the wait only actually lasts 30 minutes.

When clocks go forward 1 hour, we will expire alarms and waits up to 1 hour too early. When clocks go back 1 hour, we will expire them up to 1 hour too late.


Issue Summary: WebSphere Cape Clear profile created using the Cape Clear installer does not configure HTTPS transports
ID: SVR-2351
Component: Installer
Description: When selecting to create a WebSphere profile during the Cape Clear Server installation, the installer creates developer profiles for WebSphere 6.1. In these profiles, the HTTPS transports are not configured. If you wish to use HTTPS, you have to configure this transport manually.


Issue Summary: Supercede recovery policy
ID: SVR-2115
Component: Runtime -BPEL Runtime
Description: The observed behavior for the BPEL 'supercede' recovery policy for BPEL process instances on standalone servers is currently equivalent to the 'no-recover' setting. Setting a supercede value will not achieve the desired effect.


Issue Summary: Server can open too many receiver channels on IBM Websphere MQ
ID: SVR-2337
Component: Transports - JMS
Description: When reading a high number of JMS messages, Cape Clear Server may cache too many JMS Sessions and MessageConsumers. The default value for the JMS transport property "capeconnect.jms.server.inboundCapacity" should be reduced from 1000 to a lower value such as 10 in order to limit the number of JMS sessions created by the transport.


Issue Summary: Faults returned to JMS transport are discarded if jms.drop.noreplyto is set to true
ID: SVR-2005
Component: Transports - JMS
Description: If an exception occurs during the processing of a request message, the JMS transport will silently drop the returned Fault if JMSReplyTo is not set and if the jms.drop.noreplyto configuration property is set to true in the app descriptor.

Fixed Issues for Cape Clear Server 7.0.1

Issue Summary: Send and Receive takes over 5 seconds when using Websphere MQ
ID: SVR-2311
Component: Transports - JMS
Description: When working with MessageListeners, sending and receiving of messages in WebSphere MQ takes over 5 seconds. The fix is to use plain JMS receives instead.


Issue Summary: Support for WS-Security credentials propagation in BPEL required
ID: SVR-2280
Component: Security - WS-Security
Description: Requirement to propagate Username/Token information from incoming requests to outgoing partners.


Issue Summary: NullPointerException reported by Assign activity
ID: SVR-2242
Component: Runtime -BPEL Runtime
Description: A literal assign reports a NullPointerException if the destination element does not exist. In other situations where the destination element does not exist, a meaningful error message is returned:

The following NullPointerException occurs:
[02-Apr-2007 14:24:15.161 IST] ERROR <DefaultProcessor> An exception occured processing a process instance [id: 99, activity: /process[1]/sequence[1]/assign[1]/copy[1], cursor activity: /process[1]/sequence[1]/assign[1]/copy[1], cursor last executed: /process[1]/sequence[1]/assign[1]].
java.lang.NullPointerException
at com.capeclear.bpel.runtime.executor.assign.Copy.preSet(Copy.java:71)
at com.capeclear.bpel.runtime.executor.CopyExecutor.execute(CopyExecutor.java:65)
at com.capeclear.bpel.runtime.cache.impl.DefaultProcessor.doWork(DefaultProcessor.java:182)
at com.capeclear.bpel.runtime.cache.Processor$ProcessorThread.run(Processor.java:58)


Issue Summary: Completed activities not always saved to the database when process instance fails.
ID: SVR-2241
Component: Runtime -BPEL Runtime
Description: Previously, when a process instance failed, some activities that completed prior to the failure were not persisted to the database and were not visible within the Orchestration Manager. This issue has been resolved in 7.0.1.


Issue Summary: Invoke resulting in transport exception causes unitialized variable exception when operation in use has a fault defined
ID: SVR-2037
Component: Runtime -BPEL Runtime
Description: A transport exception on a invoke that uses a wsdl operation with a wsdl fault defined will lead to the transport exception being masked by an UninitializedVariableException exception.


Issue Summary: When copying from an element to a type-compatable element with a different qname the BPEL runtime should create the appropriate element.
ID: SVR-2081
Component: Runtime -BPEL Runtime
Description: When assigning from one element of simpleType to another (accessed directly as a message part, or a query into a variable), the BPEL engine does not throw an error if the element names and namespaces do not match and will assign the element anyway.

This does not affect cases where the element being assigned into already exists. In this case, only the content of the element (of simpleType) is assigned.


Issue Summary: The use of properties within Copy Statements is now supported
ID: SVR-1394
Component: Runtime -BPEL Runtime
Description: The BPEL specification allows properties to be used as part of an assignment. The Cape Clear Studio BPEL Editor provides support for creating this form of assignment and the BPEL extension XPath function bpws:getVariableProperty, which is supported by the Cape Clear Server runtime, provides support for assigments specified in the following way:

<bpws:copy>
<bpws:from expression="bpws:getVariableProperty('getValueRequestMessage', 'wrapper:identifier')/>
<bpws:to variable="setValueRequestMessage" property="wrapper:identifier" />
</bpws:copy>

The BPEL runtime does not support assignments of the form:

<bpws:copy>
<bpws:from variable="getValueRequestMessage" property="wrapper:identifier"/>
<bpws:to variable="setValueRequestMessage" property="wrapper:identifier" />
</bpws:copy>


Issue Summary: Transforms importing other schemas using xsd:import
ID: SVR-2195
Component: Runtime - Transformation
Description:

When a project containing an import of a schema is deployed, the following error is displayed:

com.capeclear.capeconnect.application.TargetException:

Invocation failed for application test-tsimport for interface com.capeclear.applications.duplo.TransformServiceImpl

- Error initialising JAX-RPC service for application test-tsimport with interface com.capeclear.applications.duplo.TransformServiceImpl

- failed to load transformer: textToXml

- failed to initialize from schema data: Invalid Schema : 1206 : An error occurred while initialising the text schema processor for the element {http://schemas.capeclear.com/2004/01/TextSchema/Swift/MT54X}Transactions: [/xsd:schema[@targetNamespace=&quot;

http://schemas.capeclear.com/2004/01/TextSchema/Swift/MT54X"]/

xsd:element[@name=&quot;MT54X&quot;]/

xsd:complexType/xsd:sequence/xsd:element[1]] Invalid Schema: 13292 : Failed to find definition of the referenced element &apos;

{http://schemas.capeclear.com/2004/01/TextSchema/Swift/types}

SectionOne&apos;.

This error is similar to the Studio error where an imported type is not being found.


Issue Summary: MessageStoreInterceptor outputs ignorable error when marking fault messages delivered
ID: SVR-2212
Component: Message Store
Description: When a fault occurs during processing of a BPEL request where the fault happens before the BPEL engine receives the message and marks it delivered, the MessageStoreInterceptor may output the message

ERROR <MessageStoreInterceptor> Failed to mark request message ID=5 delivered on fault. Reason: Cannot commit/rollback as no persistence transaction exists.

The delivered status of the message is however correctly updated and this message can be ignored.


Issue Summary: Client interceptors configured in Studio not invoked at runtime
ID: SVR-2131
Component: Interceptors
Description: This problem has been observed in 6.7 and 6.9.800.4 and 6.9.800.146 so is likely to effect all recent versions. It would appear that interceptors configured by Studio in the ws-application-client.xml file are not invoked at runtime. If the SOAP interceptor is programmatically added to the client proxy, it is invoked correctly.


Issue Summary: CC 7.0 Installation Problem
ID: SVR-2240
Component: Installer
Description: Deselecting the Orchestration module caused a NoClassDefFoundError on startup of the server. The Orchestration module is not optional in 7.0.1.


Issue Summary: isInRole does not work and Subject not made available
ID: SVR-2244
Component: CC-Services
Description: Currently our role support does not work. Also, we do not expose the Subject which is the result of authentication, so it is difficult for custom code to do security handling on anything but the principal. It also means our authentication handling is completely separate from WS-Security. Providing the Subject which is the result of the JAAS login would alleviate this problem.


Issue Summary: Manager throws error when expanding category with +
ID: SVR-2224
Component: Administration - Manager
Description: Fixed a series of Javascript errors on the Cape Clear Manager admin console. For example, clicking on the + beside the Server or Adapters links resulted in a Javascript error.

Note that if you were running Manager from within Cape Clear Studio (Eclipse), this error is displayed immediately. If you were running in standalone Internet Explorer, it may not be displayed immediately, depending on your IE settings. However, an error icon is displayed in the bottom left-hand side of the status bar.