Known Issues
The following sections describe known issues, levels of support, notes, and general problems in WebLogic Server 7.0 and its service packs:
Known Issues in WebLogic Server 7.0 Service Pack 7
WebLogic Server 7.0 Service Pack 7 has the following known issues:
Core Server Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR174779
|
Requests from an existing session are directed to the primary server even after the primary server starts to shut down.
Workaround: Refesh the page after a short duration.
|
|
EJB Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR262695
|
Fully classified class names are not generated in the ejb.xml and weblogic-cmp-rdbms-jar.xml deployment descriptor files. Instead, only simple class names are generated for method param values.
Workaround: When using unloadable classes in your EJBs, manually edit the deployment descriptor files to refer to those EJBs with their fully classified names.
|
|
JTA Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR259037
|
When a transaction times out at logging time, you could experience a deadlock situation. The thread that is trying to write the transaction to the transaction log (TLOG) times out, a different thread trying to execute the rollback to clean the same transaction from the TLOG and deadlock each other.
Workaround: Decrease the value of the CheckpointIntervalSeconds attribute and move TLOGs to a directory where they will not run out of memory.
|
|
RMI Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR258722
|
WebLogic Server 6.1 has some limitations when acting as a client to WebLogic Server 7.0 SP7 over the IIOP protocol. This is because WebLogic Server 6.1 does not support class evolution, so only identical classes can be passed between WebLogic Server 6.1 and WebLogic Server 7.0 SP7.
Workaround: Be aware of this issue.
|
|
|
CR261149
|
Remote requests over IIOP that involve medium to large size Strings may be a little slower in WebLogic Server 7.0 SP7 when compared to 7.0SP6. This is due to changes in JDK 1.4.2_10, which is being used in WebLogic Server 7.0SP7.
Workaround: Pass the XX:+UseTLAB JVM option flag. This flag turns on thread-local memory allocation. This flag is enabled by default only for the server VM on SPARC.
|
|
Upgrade Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR265197
|
A License exception occurs when you start WebLogic Server after upgrading from any previous WebLogic Platform 7.0 release to WebLogic Platform 7.0 SP7 using the WebLogic Platform or WebLogic Server upgrade installer.
This problem occurs only when you use the evaluation license, and only on a HP-UX system.
Workaround: Replace the license.bea file in the BEA_HOME directory with the evaluation license file that is generated by installing WebLogic Platform or WebLogic Server 7.0 SP7 using the full installer.
|
|
Known Issues in WebLogic Server 7.0 Service Pack 6
WebLogic Server 7.0 Service Pack 6 has the following known issues:
Administration Console Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR213792
|
The help information for defining security roles and security policies for EJBs is missing from the Administration Console in this release of WebLogic Server. For information about assigning security roles and security policies to EJBs, see Securing WebLogic Resources.
|
|
|
CR215117
|
The attribute, JavaVMVendor (of weblogic.management.runtime.JVMRuntimeMBean), was added to the Administration Console server monitoring versions page for WebLogic Server 7.0 Service Pack 6.
There is currently no corresponding popup for it in the Administration Console online help. As a result, a 404 error occurs when an attempt is made to access the online help popup for this attribute.
|
|
JDBC Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR192616
|
The API getXAConnection() through the net driver is not supported in the IBM DB2 UDB type 2 JDBC 2.0 driver.
It is recommended that you use a different driver that supports the getXAConnection() API.
|
|
Plug-ins Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR217777
|
Deadlock occurs in Apache HTTP Server plug-in under load.
Problem: When using the Apache HTTP Server plug-in for distributing client connection requests to WebLogic Server instances in a cluster, clients may be unable to connect to the cluster and the following message is logged in the Apache HTTP Server plug-in:
[error] server reached MaxClients setting, consider raising the MaxClients setting
This problem occurs under load.
Platform: All
Workaround: When configuring the Apache HTTP Server plug-in, set the value for the MaxClients parameter in the httpd.conf configuration file to be equivalent to the number of client connections that the cluster needs to be able to support.
For more information about configuring the Apache HTTP Server plug-in, see "Installing and Configuring the Apache HTTP Server Plug-In" in Using Web Server Plug-Ins with WebLogic Server at the following URL:
http://e-docs.bea.com/wls/docs81/plugins/apache.html
|
|
Known Issues in WebLogic Server 7.0 Service Pack 5
WebLogic Server 7.0 Service Pack 5 has the following known issues:
Core Server Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR123714
|
Network classloading can fail when multiple servers, each with its own Network Interface Card, are operating in a single cluster or domain.
Failures occur because in the JVMID of the server the hostAddress is always assumed to be the default channel address. The client uses RJVMImpl.getCodeBase to locate the URL from which to download the class, tries to connect to the default channel, and fails.
In 8.1 the hostAddress of the JVMID is based on the channel the request comes from, and thus the hostAddress correctly registers the address of the channel.
|
8.1
|
|
CR079543, CR084568
|
On Unix, some server start scripts will fail if the name of the directory in which you install WebLogic Server (BEAHOME) contains a space.
|
|
Cluster Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR168598
|
In WebLogic Server 7.0, the Managed Server starts in the same mode as the Administration Server even if you specify a different mode on Managed Server startup.
For example, if you start the Administration Server in development mode and then start the Managed Server with a command line option of -Dweblogic.ProductionModeEnabled=true, the Managed Server still starts in development mode.
The Managed Server starts up in production mode only when Administration Server has been started in production mode.
|
|
|
CR135944, CR100373, CR130592, CR135864
|
Staged one-phase applications on a Managed Server in Managed Server Independence mode, that is, with the Administration Server down, do not deploy when the Managed Server is on a separate file system or machine from the Administration Server. When the Managed Server is brought up in MSI mode, a deployment error such as the following occurs:
<Feb 5, 2004 5:06:36 PM EST> <Error> <Management> <141006> <Application 474001:Name=DefaultWebApp,Type=Application not found at \home2\mks\bea70sp4\user_projects\474001\applications\DefaultWebApp><Feb 5, 2004 5:06:36 PM EST> <Error> <Management> <141004> <IOException opening application 474001:Name=DefaultWebApp,Type=Application, loading from path \home2\mks\bea70sp4\user_projects\474001\applications java.io.FileNotFoundException: File does not exist:
\home2\mks\bea70sp4\user_projects\474001\applications\DefaultWebApp at weblogic.management.mbeans.custom.Application.adminLoad(Application.java:627)
This error occurs because there is no staging mode for one-phase applications, and the file is not cached (staged) on the Managed Server.
|
8.1
|
Documentation Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR196210
|
Previous versions of this document and various other sample documents erroneously described using weblogic.management.Admin.getInstance().getAdminMBeanHome() as a way to look up the MBeanHome interface on the Administration Server.
However, the weblogic.management.Admin class is not public. Instead of using this non-public class, use JNDI to retrieve MBeanHome. See Determining the Active Domain and Servers in Programming WebLogic Server JMX Services.
|
|
EJB Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR112260
|
The <security-permission> element for an EJB was not displayed in the Administration Console, and the descriptor validation failed for the security-permission attributes. When the bean was deployed and its client called, it was possible to change the java.vm.version property.
weblogic-ejb-jar.xml contained this stanza:
<security-permission> <description>Optional explanation goes here</description> <security-permission-spec> grant { permission java.util.PropertyPermission "java.vm.vendor", "read"; } ; </security-permission-spec> </security-permission> </weblogic-ejb-jar>
Workaround: A code change was made to allow <security-permission> to be viewed and modified in the Administration Console.
To correctly configure the desired security permission, it is necessary to perform the following steps:
1. Change the grant permission in <security-permission-spec> to { permission java.util.PropertyPermission "java.vm.vendor", "read"; }
2. Add this line to the %WL_HOME%\server\lib\weblogic.policy file:
add grant codeBase "file:/<Your
user_projects dir>/YourDomain/lib/-"
{permission java.security.AllPermission;
} ;
This is required because the dynamically generated stub's
classpath is lib.
3. Start the server with the -Djava.security.manager option.
|
|
Deployment Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR099037
|
Configuring a default Web application for a cluster and then targeting or redeploying that Web application to either a single server within the cluster or to the cluster itself will result in a deployment failure.
|
|
JDBC Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR173725
|
An invalid batch value exception occurred using Oracle 10g thin driver. Analysis revealed that the exception was caused by an Oracle thin driver limit for the batch size. The limit seemed to be 16383 (half the max value for a short).
Workaround: Limit the batch size or use 920 driver.
Oracle has stated that the bug has been fixed and is now being backported on top of the 10.1.0.2 drivers.
|
|
JMS Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR108664
|
The queue browser returns an expired message even if the message had already expired at enumeration time.
|
8.1
|
Security Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR096349, CR133465
|
Please review the security advisory information at Security Advisory BEA04-43.01.
|
8.1
|
|
CR125639
|
The note in the weblogic/security/SSL/SSLSocketFactory.java file is incorrect. The note should read as follows:
This SSLSocketFactory cannot be plugged into JSSE because the getDefault() method used by JSSE returns an instance initialized to use the old Phaos SSL implementation, instead of the Certicom SSL implementation currently used by the server. The Phaos implementation does not allow you to configure the trusted certificates, while Certicom's would be configured according to the server SSL trust configuration. If you really need to access the Certicom implementation from JSSE, you can extend the javax.net.SocketFactory class and implement its getDefault() method to return an instance of your class that wraps the SocketFactory instance returned by the weblogic.security.SSL.SSLSocketFactory.getDefaultJSS() method. The following code example shows how to extend the javax.net.SocketFactory class to access the Certicom implementation from JSSE:
public class MySSLSocketFactory extends javax.net.SocketFactory {
private static MySSLSocketFactory defFactory;
private javax.net.SocketFactory factory; private MySSLSocketFactory() {
factory = weblogic.security.SSL.SSLSocketFactory. getDefaultJSSE(); }
public static javax.net.SocketFactory getDefault() {
if (defFactory == null) defFactory = new MySSLSocketFactory(); return defFactory; }
public Socket createSocket(InetAddress host, int port)
throws IOException {
return factory.createSocket(host, port); }
// implement the other three abstract createSocket
// methods the same way. ... }
|
|
WebLogic Tuxedo Connector Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR074356
|
When you configure a redirect for tBridge, the redirect configuration information appears at the wrong place in the config.xml file. The redirect configuration appears as a child of the tBridgeGlobal MBean. It should appear at the same level as the tBridgeGlobal MBean.
Workaround: Manually edit the config.xml file.
|
8.1
|
Web Services Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR091230
|
On Microsoft Windows, the clientgen Ant task may overwrite entities whose names differ only in capitalization. This occurs because Windows is case insensitive.
|
|
Known Issues in WebLogic Server 7.0 Service Pack 4
WebLogic Server 7.0 SP4 has the following known issues.
jDriver Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR094209
|
Using WebLogic Server applications with distributed transactions and Oracle Server 9.2.0.x causes problems with the Oracle thin driver 9.2.0.x.
Workaround: Use the Oracle 9.2.0.3 driver with patch 2717235. Check the Oracle Web site and apply any necessary patches.
In addition to the workaround, the following tuning technique may help alleviate the problem:
|
|
JSP Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR061104
|
When you run the JSP compiler on Windows systems, output files names are always created with lower-case names. To prevent this behavior, and preserve the case used in classnames, set the system property weblogic.jsp.windows.caseSensitive to true. You can set the property at the command line when compiling a JSP using the following command:
java -Dweblogic.jsp.windows.caseSensitive=true weblogic.jspc *.jsp
Or include this command in your WebLogic Server startup scripts:
-Dweblogic.jsp.windows.caseSensitive=true
|
|
Web Services Known Issue
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR092268
|
Using clientgen through a proxy from behind a firewall sometimes fails because the proxy user name and password are not sent.
|
|
Known Issues in WebLogic Server 7.0 Service Pack 3
WebLogic Server 7.0 SP3 has the following known issues.
Browser Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR106615
|
When using Netscape version 4.79 on HP-UX 11.0 and 11i, a browser session may fail due to bus errors.
|
|
|
CR107134
|
(JRockit) When using Oracle server version 8.1.7.3, running on Linux, and using JRockit as your JVM, creating a new EJB may result in an end-of-file error. The subsequent version of Oracle server, 8.1.7.4, does not present this problem.
|
|
|
CR108206
|
An exception thrown by libCsup.2 in HPCode is not handled by WebLogic Server jDriver code, possibly causing Weblogic Server to dump core. This problem occurs when there are three layers of code: HPCode, native code, and Java. The problem has been observed while running Oracle 9.2.0.
Workaround: Use client libraries from Oracle 9.2.0.3.
|
|
|
CR094639
|
A wildcard URL (/*) added to the security-constraint of an EJB changes the return from req.getSession(false) after a session timeout. When the security-constraint contains a single non-wildcard URL, the return from req.getSession(false) is a null. When the security-constraint contains a non-wildcard URL, AND an additional wildcard(/*) URL, the return from req.getSession(false) is a valid session even though timeout has occurred. Release fixed: WebLogic Server 7.0 SP5.
|
|
EJB Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR082058
|
If an EJB is deployed as part of an Enterprise Application, you cannot redeploy the EJB without redeploying the entire application.
|
8.1
|
|
CR096713
|
Some EJBCacheRuntime object IDs do not return results when you run the snmpwalk command against them because their RuntimeMBeans have been removed but their object IDs still exist in the Management Information Base: 150.x.x EJB Entity Home Runtime, 458.x.x EJB Stateful Home Runtime, 170.x.x EJB Stateless Home Runtime, 137.x.x EJBCache Runtime, 136.x.x EJB Cache Monitor Runtime, 150.x.x EJB Entity Home Runtime, 160.x.x EJB Message Driven Runtime, 161.x.x EJB Pool Runtime, 458.x.x EJB Stateful Home Runtime, 170.x.x EJB Stateless Home Runtime, 455.x.x Connector Connection Pool Runtime, and 312.x.x JTA Statistics Runtime.
|
|
Tools Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR092278
|
In order to redeploy a Web Application using the weblogic.Deployer utility, you must specify both the -source and -target options on the command line. If the -source and -target options are not specified, the tool indicates success but does not redeploy the Web Application.
|
|
Web Services Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR101160
|
Complex data types in service-specific exceptions are not handled in strict accordance with the Java API for XML-Based RPC (JAX-RPC) 1.0 specification, as follows:
Section 4.3.6 WSDL Fault, Service Specific Exception: When the clientgen Ant task generates the JAX-RPC stubs from a WSDL that contains a wsdl:fault element whose message part is an xsd:complexType, the generated stubs incorrectly map the xsd:complexType to a simple Java type. This means that when a client application uses the stubs to invoke the Web Service, and the service throws the service-specific exception (rather than the a javax.xml.rpc.SOAPFaultException), the client application cannot access the full complex data from the exception.
There is no workaround to this problem.
Section 5.5.5 Methods, Exceptions, fourth bullet: The WebLogic Web Service runtime incorrectly maps multiple fields in a service-specific Java exception class to a simple XML data type, rather than to an xsd:complexType.
Workaround: Always throw a javax.xml.rpc.SOAPFaultException,
rather than a service-specific exception, when implementing a WebLogic Web
Service.
|
|
Known Issues in WebLogic Server 7.0 Service Pack 2
WebLogic Server 7.0 SP2 contains the following known issues.
HPUX 11i Users Must Apply a Patch
An incorrect version of the hpux11/libweblogicunix1.sl library was bundled with the WebLogic Server 7.0 SP2 download. To obtain the correct library, HPUX 11i users must download a patch from the following location, and apply it:
http://commerce.bea.com/products/weblogicserver/suppljar.jsp
To apply the patch:
- Download the patch.
- De-archive it to $WLHOME/weblogic700/server/lib:
jar -xvf CR102314_700sp2.jar
This will replace hpux11/libweblogicunix1.sl with a new library.
- Prepend the .jar file to your CLASSPATH.
Known Issue with JDK 1.4.1 Support and SSL
In order for JDK 1.4.1 clients to use SSL, the Java 2 SDK jsse.jar must be removed from the jre\lib directory:
Directory of C:\j2sdk1.4.1\jre\lib:
08/29/2002 01:10a 462,386 jsse.jar 1 File(s) 462,386 b
Administration Console Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR110791
|
When you add custom help pages to the Administration Console, use the Console Extension described in Extending the Administration Console.
If you use the alternative method of placing custom help files in console.war, be sure to add console.war to WL_HOME/server/lib/weblogic.jar. Otherwise the custom files will be wiped out when console.war is regenerated at server startup.
|
N/A
|
|
CR066112
|
Deployment procedures using the Administration Console have changed since the 6.x release. For more information, see WebLogic Server Application Deployment.
|
N/A
|
|
CR069828
|
You can no longer edit the Web Services deployment descriptor using the Administration Console.
|
N/A
|
|
CR069927
|
Selecting Server --> Monitoring --> Process Output and selecting any of the view options does not display the output. If you select the "View Node Manager Output" while Node Manager is running, Node Manager will crash.
|
7.0 SP3
|
|
CR073164
|
You cannot use the Administration Console to edit a domain other than the current domain of the Administration Server.
|
|
|
CR074339
|
Explanatory text for creating configuration objects is not displayed when using Netscape 4.7 to access the console.
|
8.1
|
|
CR094174
|
The Administration Console gives a RuntimeError with Internet Explorer version 5 when you select JDBC->MultiPool in the left panel of the Console.
|
|
Core Server Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR048834
|
On Solaris, if you are using JDK 1.3.1, and if the timestamp for WebLogic log messages is in Greenwich Mean Time (GMT) regardless of your local time settings, you need to update jre/lib/i18n.jar.
In some cases, this JAR file is missing the resources that the JRE uses to localize time stamps.
Workarounds:
|
|
|
CR059543
|
If you use the Administration Console to update an XML Registry by setting the DocumentBuilderFactory text field to NULL, then subsequently view the same XML Registry, the DocumentBuilderFactory text field is empty. It should contain WebLogic Server's default DocumentBuilderFactory, weblogic.apache.xerces.jaxp.DocumentBuilderFactoryImpl.
|
|
|
CR063799
|
Calling getLogFileName on VirtualHostMBean returns an incorrect path.
|
8.1
|
|
CR065299, CR067642
|
The Java HotSpot(TM) Server VM is not yet supported on WebLogic Server 7.0. Use the certified client version of this VM.
|
|
|
CR066581
|
The WebLogic Server version of javax.management.timer.Timer, called weblogic.management.timer.Timer, was incompatible with the JMX timer.
If you use the WebLogic class and instantiate it, you will have timers that use the WebLogic thread pools and honor the user identity when executing their listeners.
|
|
|
CR066723
|
Managed Server discovery is off by default when you are in development mode. To turn it on, add -Dweblogic.manager.discover=false to your command line argument that starts WebLogic Server.
|
|
|
CR067087
|
WebLogic Server currently does not provide a means to set security permissions within a weblogic-applications.xml file.
|
|
|
CR067814
|
If you wish to use an LDAP browser to view the embedded LDAP server on WebLogic Server, try using the following input:
host: localhost port: 7003 Anonymousbind LDAP Version 3
|
|
|
CR067868
|
When using the HotSpot JVM the server can hang when running large deployments.
Workaround: Try setting your -XX:MaxPermSize to a higher figure.
|
|
|
CR068186
|
If you install WebLogic Server version 7.0 on a multi-homed, Windows 2000 system, configured servers will incorrectly resolve each IP address to the common DNS name for the system itself. For example, consider a multi-homed Windows 2000 machine named "MyWin2KMachine" and having IP addresses:
192.168.1.105 192.168.1.106 192.168.1.107
If you install multiple WebLogic Server instances on the system using IP addresses 192.168.105, 192.168.1.106, and 192.168.1.107 for the servers' Listen Address attributes, all servers will yield the common DNS name, "MyWin2KMachine," when resolving their IP addresses. This can lead to conflicts when accessing resources on remote servers.
Workarounds:
192.168.1.105 MyWin2KMachine1 192.168.1.106 MyWin2KMachine2 192.168.1.107 MyWin2KMachine3
If the machine's IP addresses have names registered with a DNS server, use those externally-available DNS names instead of IP addresses when configuring servers' Listen Address attributes. For example, configure servers using MyWin2KMachine1, MyWin2KMachine2, and MyWin2KMachine3 rather than using 192.168.105, 192.168.106, and 192.168.107. Note that this problem does not affect other software platforms, including Windows NT.
|
|
|
CR068228
|
When you are not running in compatibility mode and you use the command line Administration tool, you may be prompted for a `guest' password. The default password for "guest" is "guest".
|
|
|
CR068247
|
The JSPs included with WebLogic Server 7.0 are not being pre-compiled before they are accessed for the first time. If you are deploying precompiled JSPs, WebLogic Server may recompile them to the /wlnotdelete/ directory.
|
|
|
CR068361
|
If you are migrating a 1.1 EJB from Bean Managed Persistence to Container Managed Persistence, selecting the bean and checking "Container-managed persistence" should generate Advanced and Automatic Key Generation options, but does not.
|
|
|
CR069278
|
When using the command line to create a connection pool, in the properties string you can only escape the double quotes on platforms where the shell removes one set of quotes.
Workaround: On platforms where the quotes are not removed during this process, you should use only one set of double quotes. You can also repeat the property argument for each connection property, which eliminates the need for quoting a string of properties. For example:
-property Property user=scott -property Password tiger -property URL jdbc:weblogic:oracle:ora817
|
|
|
CR069602
|
The JTA implementation in WebLogic Server was updated to comply with the JTA specification version 1.0.1a, including minor changes to the Transaction.commit() and TransactionManager.setTransactionTimeout(int) methods.
|
|
|
CR072646
|
When using Netscape version 4.75 or older on Windows 2000, file associations may not be converted correctly from the same used for Internet Explorer. This results in the following behavior when opening About WebLogic Server 7.0, About Examples, and About Pet Store from the Windows Start menu:
|
|
|
CR072940
|
If your config.xml file does not have the <SecurityConfiguration> element set, or is missing the credential and name attributes within that element, your server tries to create this attribute. If your config.xml file is writeable, your java client caches the value of this credential in its local JVM memory space. Then, for the lifetime of the JVM, the java client uses this credential. However, once the JVM gets re-started, it is denied access because it does not have the credential.
Workaround: save your config.xml file once, when it has the credential already set.
|
|
|
CR073395
|
There is a problem with Sun's Pet Store application when you try to run it with more than one client and use an Oracle database.
|
|
|
CR073647
|
Currently only users in an Administration role can use the AdministrationPort.
|
8.1
|
|
CR073737
|
If you restart an Administration Server for an active domain, the Administration Server can fail to discover its Managed Servers. This occurs with the following configuration:
A multihomed machine running a Windows operating system, in which the Administration Server and one or more processes use the same listen ports (but on different IP addresses). It is caused by a limitation of the Windows operating system. If a restarted Administration Server does not discover its Managed Server, commands that you issue to a domain via the Administration Server will fail. For example, a Managed Server that starts in Managed Server Independence (MSI) mode can fail to respond to shutdown commands after the Administration Server regains control.
|
|
|
CR073793
|
You cannot deploy an application with untargeted modules if you have not set URIs for the untargeted modules in your config.xml.
|
8.1
|
|
CR074296
|
In some situations, the Administration Console allows users to create more than one WebLogic Tuxedo Connector WTCResources MBean. Only one WTCResources MBean is allowed for a given WTCServer MBean configuration. Users may need to remove the extra configuration and restart the WebLogic Tuxedo Connector.
|
8.1
|
|
CR074452
|
If your JVM is throwing errors and WebLogic Server is under heavy load, try increasing your java memory heap.For example, modify the command that starts WebLogic Server by adding the following: MEM_ARGS="-XX:MaxPermSize=128m -Xms512m -Xmx512m"
|
|
|
CR074969
|
You may encounter problems deploying EJBs without JNDI names. BEA recommends that users continue to deploy their EJBs with JNDI names as was done in earlier releases of WebLogic Server.
|
|
|
CR085169
|
When using Iplanet 4.1SP9, the Java client can not read post data from a plug-in due to read_timeout from the java client using socket.getInputStream().readLine()' caused by a problem in netbuf_getbytes().
Workaround: Using Iplanet 6.0SP2 with WebLogic Server eliminates this problem.
|
|
|
CR099314
|
There is a problem with JDK 1.3.1_04 and the nohup command, as described at http://developer.java.sun.com/developer/bugParade/bugs/4755829.html.
The nohup command starts failing; the WebLogic Server process exits when the parent shell exits, regardless of the invocation of nohup. The problem was confirmed for ksh (the default Solaris 8 shell), but may exist for other shells. The problem does not occur with the bash shell.
Workarounds:
|
|
|
CR110028, CR120181
|
Stuck thread detection in WebLogic Server 7.0 is limited to threads in the default execute queue.
Stuck thread detection is done by checkStuckThreads method of weblogic.t3.srvr.CoreHealthService, and a private variable, policyName, which corresponds to the queue name for checking stuck threads, is statically set to "default."
|
|
EJB Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR089847
|
Many WebLogic Server EJB deployment properties have default values that are optimized for performance. In some cases, these default values are not compliant with the EJB specification.
Workaround: To make WebLogic Server compliant with the EJB specification, set the following properties:
In weblogic-ejb-jar.xml set enable-call-by-reference to False.
In weblogic-cmp-jar.xml, set include-updates to True.
In weblogic-cmp-jar.xml, set check-exists-on-method to True.
|
N/A
|
|
CR089630
|
When a table contains both BLOB and LONG RAW columns, OracleThinDriver requires that the LONG RAW column be read first from the stream even if BLOB occurs before LONG RAW in the select list.
The EJB container currently does not handle this case and throws a java.sql.SQLException "Stream already closed".
Workaround: Use JDriver.
|
|
|
CR097323
|
There is a known issue with deploying or redeploying EJBs to a single server instance in a cluster—referred to as pinned deployment—if the .jar file contains contain uncompiled classes and interfaces.
During deployment, the uncompiled EJB is copied to each server instance in the cluster, but it is compiled only on the server instance to which it is deployed. As a result, the server instances in the cluster to which the EJB is not targeted lack the classes generated during compilation that are necessary to invoke the EJB. When a client on another server instance tries to invoke the pinned EJB, it fails, and an Assertion error is thrown in the RMI layer.
Workaround: If you are deploying or redeploying an EJB to a single server instance in a cluster, compile the EJB before deploying it to ensure that the generated classes are copied to all server instances and available to all nodes in the cluster.
|
N/A
|
Examples and Pet Store Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR055626
|
The login.jsp file for the webapp security sample does not allow for URL encoding when a user logs in. The login.jsp file is in this location:
SAMPLES_HOME\server\src\examples\webapp\security
The login.jsp file has this code:
<form method="POST" action="j_security_check">
Here is an example of code that could be used to allow for URL encoding:
<FORM ACTION="<%= response.encodeURL("j_security_check") %>" METHOD=POST>
|
7.0 SP2
|
|
CR068009
|
Due to Microsoft VB Scripting Versioning, Excel 2000 or later is required to run the ZeroClient jCom example.
|
|
|
CR071260
|
The password and username for the examples and Pet Store has been changed since WebLogic Server 7.0 Beta. Using installadministrator/installadministrator no longer works, use weblogic and weblogic instead.
|
N/A
|
|
CR073395
|
Pet Store has problems when multiple clients are using the same resources. Sun is working to address this problem.
|
|
|
CR074004
|
In the package-summary.html file in the SAMPLES_HOME\server\src\examples\security directory, the Security Examples Summary table makes references to the audit and cert samples. These samples have been removed from the WebLogic Server distribution because they used APIs that are being deprecated in WebLogic Server 7.0.
Also, the descriptions of the SSLClient, JAAS, and Net samples should read as follows:
SSLClient: Demonstrates how to make outbound SSL connections and two-way SSL connections from a WebLogic Server that is acting as a client to another WebLogic Server or application server.
JAAS: Demonstrates use of the Java Authentication and Authorization Service by validating username and password, and then, if valid, executing the stateless session Trader EJB.
Net: Demonstrates writing connection filters that implement the weblogic.security.net.ConnectionFilter and ConnectionFilterRulesListener interfaces. Connection filters allow you to reject or accept client connections to the WebLogic Server environment based on the client's IP address and protocol and the local and remote port numbers.
|
N/A
|
|
CR080288
|
There are problems with the RMI cluster example when using ant run to execute the client. The java call to the client includes incorrect quotation marks.
|
|
Installation Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR085531
|
After successful upgrades and downgrades of a WebLogic Server via SmartUpdate, a subsequent upgrade results in the following error:
Installer is not compatible with the installed product. Your product directory contains installed components that are not available in the current installer. Please check your installer to ensure it matches the product distribution on your system.
Workaround: Contact your customer support representative for assistance.
|
|
Interoperability Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR093343 and CR093347
|
When a Java client running on Java 2 SDK 1.4.1 accesses a server running on Java 2 SDK 1.3.1, the following errors occur:
These are known issues with the Sun Java 2 SDK (Ticket #808980). Sun plans to fix these errors in Java 2 SDK 1.5.
Workaround: Run the client application on SDK 1.3.1.
|
|
|
CR092507, CR095479
|
Remote methods throwing MarshalException/UnmarshalException when interoperating between 7.0 SP1 and 7.0 SP2 servers may result in unmarshalling problems. This occurs because MarshalException and UnmarshalException map to the correct completion statuses in SP2 but the incorrect completion statuses in SP1.
|
|
|
CR095804
|
When a WebLogic Server 7.0 SP2 client attempts to get the Coordinator remote object from a WebLogic 8.1 server instance, a ClassCastException is thrown while casting the remote stub to the CoordinatorOneway interface. This prevents commit handoff to the 8.1 coordinating server.
|
7.0 SP3
|
jDriver Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR094209
|
Using WebLogic Server applications with distributed transactions with Oracle Server 9.2.0.x causes problems with the Oracle thin driver 9.2.0.x.
Workaround: Use the Oracle 9.2.0.3 driver with patch 2717235. Check the Oracle Web site and apply any necessary patches.
In addition to the workaround, the following tuning technique may help alleviate the problem:
|
|
JMS Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR075438
|
A message-driven bean (MDB) that is deployed on a distributed topic, and which is targeted to a server instance that is hosting two members of the distributed topic, is also deployed on all the members of the distributed topic.
This is the expected behavior because the MDB is pinned to the member destination name. Therefore, if a JMS server contains two members, there will be two MDBs deployed, one for each member.
|
|
JRockit Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR093270
|
When running WebLogic Server 7.0 SP2 server on the JRockit JVM, in UDDI, publishing a new Web service to the private registry fails with this error message: "An error has occurred E_fatalError(10500): a serious technical error has occurred while processing the request."
This is due to a bug in the UDDI implementation (AUDDI 1.0).
|
|
JTA Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR089804
|
In a transaction, you cannot use connections from two Tx Data Sources that point to the same connection pool. When you attempt to get a connection from the second Tx Data Source, you will get an XA_PROTO error.
Internally, WebLogic Server uses the connection pool name to identify the resource, not the data source name. When you get two separate connections from the same connection pool using different Tx Data Sources, they are treated as the same resource. In a transaction, start/enlist is called on the first Tx Data Source, then start/enlist is attempted on the second Tx Data Source. Because the server considers the resources to be the same, an XA_PROTO error is thrown because end/delist was not called before starting a new transaction.
Workarounds:
There should be no performance degradation from either of these workarounds.
|
|
Node Manager Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR074230
|
Applications deployed to WebLogic Server should perform file lookups relative to the Root Directory obtained with the ServerMBean.getRootDirectory() method (this defaults to the "." directory). For example, to perform a file lookup, use code similar to:
String rootDir = ServerMBean.getRootDirectory() //application root directory
File f = new File(rootDir + File.separator + "foo.in");
rather than simply:
File f = new File("foo.in");
If an application is deployed to a server that is started using Node Manager, use the following method calls instead:
String rootDir //application root directory
if ((rootDir = ServerMBean.getRootDirectory()) == null) rootDir = ServerStartMBean.getRootDirectory(); File f = new File(rootDir + File.separator + "foo.in");
The ServerStartMBean.getRootDirectory() method obtains the Root Directory value that you specified when configuring the server for startup using Node Manager. (This corresponds to the Root Directory attribute specified the Configuration->Remote Start page of the Administration Console.)
A Managed Server started remotely by Node Manager does not recognize the weblogic.RootDirectory set on the Remote Start tab.
To avoid this problem, enter an absolute path for Staging Directory on the Domain—>Servers—>Configuration—>Deployment tab. Do not enter a relative path. Enter the full path, including the WebLogic Server root directory.
|
|
|
CR073078
|
When the Administration Server attempts to set up an SSL connection to Node Manager, it does not automatically search for certificate files in the Administration Server's root directory. If you store SSL certificate files in the Administration Server's root directory, make sure you configure the absolute path to the certificate files using the Connections->SSL tab in the Administration Console.
|
8.1
|
|
CR075142
|
The Administration Server and Node Manager may simultaneously attempt to close a TCP socket, causing the following error message to appear in the Node Manager log file:
<Apr 18, 2002 5:32:55 PM EDT> <Error> <NodeManager@localhost:5555> <OutputHandler: done2 socket close failed, reason: java.net.SocketException: Connection reset by peer.>
This problem does not affect the operation of Node Manager or the Administration Server.
|
|
|
CR094722
|
When Managed Servers in a cluster are booted, a socketWrite error is thrown on the Administration Server.
Workaround: Ignore the problem or restart the managed node.
|
|
Plug-in Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR095161
|
Under load, an Apache plug-in may receive a CONNECTION REFUSED error from a back-end WebLogic Server instance. This error was observed with Apache Server version 1.3.22 with the Apache plug-in and WebLogic Server running on Solaris 8.
|
|
RMI Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR099693
|
A client accessing WebLogic Server through a firewall using the t3 protocol resulted in an error because the server expected to route a message over an uninitialized connection.
The server's ExternalDNSName was specified. Upon initial connection, the client addressed the request to the server instance's ExternalDNSName. The JVMID in the response from the server instance required the server's internal IP address and internal DNS name. Subsequent requests from the client were not recognized.
The problem occurs if clients are connecting over t3 through a firewall. The workaround is to add an entry to the external DNS servers that maps the server instance's internal DNS name to its external IP address.
|
7.0 SP03
|
Security Known Issues
For information about BEA security advisories, refer to the BEA Advisories & Notifications page on the dev2dev Web site. On this page, you can download security-related patches and register to receive notifications of newly available security advisories. You can access the Advisories & Notifications page with the following URL: http://dev2dev.bea.com/advisories.
BEA has established an e-mail address (security-report@bea.com) to which you can send reports of any possible security issues in BEA products.
The following table summarizes security-related known issues for this service pack release:
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR086158
|
Please review the security advisory information at http://dev2dev.bea.com/resourcelibrary/advisoriesnotifications/BEA02-22.jsp.
|
|
|
CR092546
|
The proper syntax for a role name must be an XML_NMTOKEN as defined in the specification at:
http://www.w3.org/TR/REC-xml#NT-Nmtoken
|
|
|
CR093652
|
In order for clients built with JDK 1.4.1 to use SSL, the Java 2 SDK jsse.jar file must be removed from the JAVA_HOME\jre\lib directory.
|
|
|
CR094566
|
To use the nCipher JCE provider with this release of WebLogic Server, clients must be built using JDK 1.3.1.
|
|
|
CR096589
|
weblogic.servlet.security.ServletAuthentication.strong() is broken and fails with a class cast exception.
Workaround: Install the following patch: CR096589_70sp2.jar.
|
7.0 SP3
|
Third-Party Tools Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR071527
|
There is a problem persisting sessions with Resonate—a third-party load balancing tool that uses WebLogic Server as a backend server. Because Resonate does not handle case insensitivity, if URLEncoding sends a cookie as jsessionid instead of JSESSIONID, Resonate does not detect the cookie.
|
|
Tools Known Issues
|
Change Request Number
|
Description
|
Release Fixed
|
|
CR064893
|
The following deployment descriptor elements cannot be edited in WebLogic Builder:
ejb-jar.xml elements:
weblogic-cmp.xml elements:
|
|
|
CR065250
|
WebLogic Builder may fail to access a database on a server if the driver classes for the database are not on the client.
|
|
|
CR066466
|
WebLogic Builder does not provide the option to cancel operations.
|
|
|
CR066668
|
On Linux, WebLogic Builder's file browser opens at root instead of current directory.
|
|
|
CR068152
|
The operations Add, Edit, and Delete can be slow in the EJB elements of WebLogic Builder.
|
8.1
|
|
CR068165
|
In the Add dialog in Resource Links, in the WebLogic Settings, J2EE Links, WebLogic Builder incorrectly offers a selection of application scoped data resources.
|
|
|
CR068620
|
Table columns may not resize properly in WebLogic Builder.
|
8.1
|
|
CR069568
|
WebLogic Builder cannot map multiple roles to a method in EJB method permission.
|
|
|
CR071833, CR072541
|
WebLogic Builder support for RARs does not include generation of deployment descriptors.
|
|
|
CR072700, CR074591
|
If you use WebLogic Builder's deployment dialog box to undeploy an application, the dialog will not show the undeployment. This dialog may also report that applications whose deployment has failed are deployed.
|
7.0 SP3
|
|
CR073975
|
WebLogic Builder incorrectly allows you to add a transaction without specifying a method or a transaction attribute.
|
7.0 SP3
|
|
CR074210
|
Optimistic concurrency strategy does not work in WebLogic Builder.
|
8.1
|
|
CR074322, CR074856
|
In WebLogic Builder, canceling an edit action may not work.
|
8.1
|
|
CR074566
|
Finders or container-managed persistence values modified through WebLogic Builder's BeanGrid/Edit (like include-update) may not be represented in the individual panels.
|
8.1
|
|
CR074622
|
In editing a container-man |