Oracle® Application Server Administrator's Guide
10g Release 2 (10.1.2) B13995-06 |
|
Previous |
Next |
This appendix provides information on how to troubleshoot problems that you might encounter when using Oracle Application Server. It contains the following topics:
See Also:
|
Oracle Application Server components generate log files containing messages that record all types of events, including startup and shutdown information, errors, warning messages, access information on HTTP requests, and additional information. The log files can be used to identify and diagnose problems. See Chapter 5, "Managing Log Files" for more information about log files.
This section describes common problems and solutions. It contains the following topics:
Oracle Application Server Infrastructure Instance Will Not Start
Application Performance Impacted by Garbage Collection Pauses
OracleAS Web Cache Fails to Initialize or Restart a Managed Process
Failure to Bring Up Standby Instances After Failover or Switchover
Previously Working Application Using ADF Business Components Starts Throwing JDBC Errors
The Oracle Application Server Infrastructure will not start.
Problem
Some common symptoms and likely causes of this problem are:
opmnctl
startall/stopall is in a hung state. This may indicate that the listener or database is down.
OPMN startup fails or is in an unstable state and opmnctl
startall/stopall is in a hung state. This may indicate that the machine has run out of memory.
You are receiving a missing component error. This usually indicates that entries are missing in the opmn.xml file or someone has incorrectly edited the file.
Oracle Internet Dependency failed. This occurs when the correct order of starting and stopping is not followed.
Solutions
The following list provides solutions to problems in the same order as the symptoms listed earlier:
Ensure that the database and listener are running.
Ensure that the machine's memory meets the memory requirements listed in the Oracle Application Server Installation Guide.
Check the opmn.xml file to see if the missing component is correctly entered in the file.
The correct order in which to start Infrastructure components is:
Oracle Database Server Net Listener
Metadata Repository
Identity Management
Application Server Control Console
Note: See Section 3.2.3, "Starting OracleAS Infrastructure" for instructions on how to start an Infrastructure. |
For information on resetting the ias_admin
password, see Section J.3.1.1, "Resetting the Administrator (ias_admin) Password".
For information on restoring a backup to a different host, see Section 22.2.3, "Restoring an Infrastructure to a New Host".
Application performance slows or application is unresponsive.
See the section "Application Performance Impacted by Garbage Collection Pauses" in the Oracle Application Server Containers for J2EE User's Guide for information about the causes and solutions to this problem.
In load conditions (for example, when the number of users concurrently connecting to the application server increases significantly in a short time), the server may respond with the following error message:
IOException in sending request - Connection refused
Problem
If the number of simultaneous users is increasing, the server may be utilizing the maximum Oracle HTTP Server child processes allowable to service requests.
Solution
You may need to increase the MaxClients directive for the Oracle HTTP Server. The MaxClients directive specifies a limit on the number of clients who can simultaneously connect.
Use one of the following methods to determine if this is the case:
Search the Oracle HTTP Server error log file for the following message:
server reached MaxClients setting, consider raising the MaxClients setting ...
By default, the error log file is located in:
(UNIX) ORACLE_HOME/Apache/Apache/logs/error_log (Windows) ORACLE_HOME\Apache\Apache\logs\error_log
Interactively monitor child process activity using the metrics in the Application Server Control Console. In particular, view the following in the HTTP_Server Home page:
In the Status section, Active Connections, which shows the number of clients currently executing HTTP requests.
In the Response and Load section, Active Requests, which shows the total number of active requests currently being processed.
Look at the information provided by mod_status. The mod_status module provides an HTML page that shows the current server statistics. Check to see if all the processes are busy. (By default, Mod_status is enabled for localhost access only.) For more information on mod_status, see:
In addition, consider increasing the maximum queue length for pending connections (the ListenBackLog directive) and consider the impact of persistent connections (the KeepAlive directive).
For more information about the Oracle HTTP Server directives and how to change their values, see the Oracle HTTP Server Administrator's Guide. For more information about tuning Oracle HTTP Server processes, see the Oracle Application Server Performance Guide.
You can get the following error if Oracle HTTP Server is unable to start due to port conflict:
[crit] (98) Address already in use: make_sock: could not bind to port 7778
See the section "Oracle HTTP Server Unable to Start Due to Port Conflict" in the Troubleshooting Oracle HTTP Server appendix of the Oracle HTTP Server Administrator's Guide for information about the cause and solution to this problem.
When there are too many httpd processes running on a machine, the response time plummets.
See the section "Machine Overloaded by Number of HTTPD Processes" in the Troubleshooting Oracle HTTP Server appendix of the Oracle HTTP Server Administrator's Guide for information about the cause and solution to this problem.
You are not able to start an Oracle Application Server process using OPMN.
See Section A.1.1, "Oracle Application Server Process Does Not Start" in the Oracle Process Manager and Notification Server Administrator's Guide for information about the causes and solutions to this problem.
On some computers, when OPMN starts up, it consumes large amounts of CPU processing capability.
See Section A.1.15, "OPMN Start Up Consumes CPU Processing Capability" in the Oracle Process Manager and Notification Server Administrator's Guide for information about the causes and solutions to this problem.
OPMN cannot start. This may be caused by a corrupt opmn.xml
file.
See Section B.1.1, "OPMN Cannot Start" in the Distributed Configuration Management Administrator's Guide for information about the causes and solutions to this problem.
The DCM daemon does not start.
See Section B.1.2, "DCM Daemon Cannot Start " in the Distributed Configuration Management Administrator's Guide for information about the causes and solutions to this problem.
DCM returns ADMN-100999 and the base exception is "Unable to connect to Directory."
See Section B.1.8, "Unable to Connect to the Directory" in the Distributed Configuration Management Administrator's Guide for information about the causes and solutions to this problem.
DCM returns ADMN-202026 and the base exception is "Unable to connect to Directory."
See Section B.1.9, "Cannot Access the Infrastructure Database" in the Distributed Configuration Management Administrator's Guide for information about the causes and solutions to this problem.
OracleAS Web Cache fails to initialize or restart a managed process.
Problem
You might receive, in the OracleAS Web Cache event log, the following errors when OracleAS Web Cache fails to initialize or fails to restart a managed process:
Error Restarting Web Cache. Reason Web Cache failed to restart a managed process after the maximum retry limit
The following errors:
[25/Nov/2004:19:12:40 +0000] [notification 9403] [ecid: -] Maximum number of file/socket descriptors set to 950. [25/Nov/2004:19:12:40 +0000] [notification 13002] [ecid: -] Maximum allowed incoming connections are 700 [25/Nov/2004:19:12:40 +0000] [alert 13305] [ecid: -] Failed to assign port 7777: Address already in use [25/Nov/2004:19:12:40 +0000] [alert 9707] [ecid: -] Failed to start the server. [25/Nov/2004:19:12:40 +0000] [alert 9609] [ecid: -] The server process could not initialize. [25/Nov/2004:19:12:40 +0000] [notification 9610] [ecid: -] The server is exiting.
Solution
Check to see if the assigned port number is occupied by executing the following command:
netstat -a |grep "port number"
Also, check to see if the port number is less than 1024. If the port number is less than 1024, but it can be set to a higher number, set the port number to a number greater than 1024.
If the listen port number must be set to less than 1024 (typical setting for SSL listen ports), then, on UNIX, you must enable the Oracle Application Server instance as root. For instructions on setting the port number to less than 1024, please refer to Section 4.3.5.1, "Changing the OracleAS Web Cache Listen Ports".
Browsers return an error saying that a page cannot be displayed.
See Section E.1.7, "Browser Displaying a Page Not Displayed Error" in the Oracle Application Server Web Cache Administrator's Guide for information about the causes and solutions to this problem.
You are not able to access your portal instance. For example, pages are not displayed, or you get an "HTTP 503 Service Unavailable" error when you try to access OracleAS Portal.
See Section K.1.1, "Unable to Access OracleAS Portal" in the Oracle Application Server Portal Configuration Guide for information about the causes and solutions to this problem.
You can access the public home page but are unable to log in. Common symptoms of this problem are the following:
The login page does not appear after you click Login.
You get an error after you enter your credentials on the OracleAS Single Sign-On login page.
You get errors on OracleAS Portal pages after you have been authenticated.
See Section K.1.2, "Unable to Log In to OracleAS Portal" in the Oracle Application Server Portal Configuration Guide for information about the causes and solutions to this problem.
Either oidctl
or opmnctl
fails to start an Oracle Internet Directory server instance.
See Section J.1.11.1, "Oracle Internet Directory Server Does Not Start" in the Oracle Internet Directory Administrator's Guide for information about the causes and solutions to this problem.
LDAP search performance is poor.
See Section J.1.5.1, "Poor LDAP Performance" in the Oracle Internet Directory Administrator's Guide for information about the causes and solutions to this problem.
Users may see an Authentication Failed error after logging in to OracleAS Single Sign-On.
See Section A.1.4, "Authentication Failed" in the Oracle Application Server Single Sign-On Administrator's Guide for information about the causes and solutions to this problem.
Logging into OracleAS Single Sign-On might take a long time depending on your configuration.
See Section A.2.1, "Logging into OracleAS Single Sign-On" in the Oracle Application Server High Availability Guide for information about the causes and solutions to this problem.
In the OracleAS Disaster Recovery standby site, you may find that the site's OracleAS Metadata Repository is not synchronized with the OracleAS Metadata Repository in the primary site.
See Section A.3.1, "Standby Site Not Synchronized" in the Oracle Application Server High Availability Guide for information about the causes and solutions to this problem.
Standby instances are not started after a failover or switchover operation.
See Section A.3.2, "Failure to Bring Up Standby Instances After Failover or Switchover" in the Oracle Application Server High Availability Guide for information about the causes and solutions to this problem.
For information about diagnosing FRM-xxxxx errors from OracleAS Forms Services, see Section A.2, "Diagnosing FRM-XXXXX Errors" in theOracle Application Server Forms Services Deployment Guide.
For information about resolving memory problems for OracleAS Forms Services, see Section A.6, "Resolving Memory Problems" in theOracle Application Server Forms Services Deployment Guide.
When running report requests with the OracleAS Reports Server, the report request may "hang" for various reasons. This can lead to stability issues if not noticed in time.
See the section "Hanging Report Requests" in Oracle Application Server Reports Services Publishing Reports to the Web for information about the causes and solutions to this problem.
A list of values (LOV) might be greater than the URL limit.
See Section D.1.14, "List of Values (LOV) Is Too Long for a Discoverer Portlet URL" in the Oracle Business Intelligence Discoverer Configuration Guide for information about the causes and solutions to this problem.
Discoverer end users might encounter errors if the OC4J_BI_forms JVM process runs out of memory.
See Section D.1.18, "Out of memory problems for the OC4J_BI_forms JVM process" in the Oracle Business Intelligence Discoverer Configuration Guide for information about the causes and solutions to this problem.
If Discoverer Portlet Provider is not correctly registered in OracleAS Portal, you mightencounter errors when creating or editing Discoverer portlets.
See Section D.1.20, "Discoverer Portlet Provider issue" in the Oracle Business Intelligence Discoverer Configuration Guide for information about the causes and solutions to this problem.
An application that previously successfully retrieved data suddenly starts throwing JDBC errors such as Connection Reset By Peer, Connection Closed, or Socket Reset By Peer.
See Section B.1.1, "Previously Working Application Using ADF Business Components Starts Throwing JDBC Errors" in the Oracle Application Development Framework Development Guidelines Manual for information about the causes and solutions to this problem.
This section describes problems that you might encounter when using Application Server Control and explains how to solve them. It contains the following topics:
See Also: Chapter 23, "Troubleshooting the Backup and Recovery Tool" for information about any troubleshooting backup and recovery operations performed within the Application Server Control Console |
This section describes problems and solutions. It contains the following topics:
Unavailable Metric and Chart Data in the Application Server Control Console
Application Server Status Is Down When Server Components Are Up
Problems Connecting to an Application Server Instance from Farm or Cluster Page
Application Server Home Page Indicates That the Farm Is Unavailable
SSL Timeout Issues with Microsoft Internet Explorer Browsers
Session Has Expired Message When Using Multiple Browser Windows
No Propagation Between Grid Control and Application Server Control When Creating a New OC4J Instance
Problems Viewing Metrics When Configured for Secure Sockets Layer (SSL)
Problems Displaying the Date Selection Window When Searching the Log Repository
To manage an instance of Oracle Application Server, you must log in to the Application Server Control Console using the current Administrator (ias_admin
) password.
Problem
If you forget or do not know the ias_admin
password, then you cannot monitor or administer the application server or its components with the Application Server Control Console.
Solution
Reset the ias_admin
password using the following procedure while you are logged in as the user who installed the Oracle Application Server instance:
Stop the Application Server Control.
On UNIX systems, enter the following command in the Oracle home of the application server instance:
ORACLE_HOME/bin/emctl stop iasconsole
On Windows systems, use the Services control panel to stop the Application Server Control service.
Locate and open the following file in a text editor:
ORACLE_HOME/sysman/j2ee/config/jazn-data.xml
Locate the line that defines the credentials property for the ias_admin
user.
The following example shows the section of jazn-data.xml
with the encrypted credentials
entry in boldface type:
<realm>
<name>enterprise-manager</name>
<users>
<user>
<name>ias_admin</name>
<credentials>{903}buG0lUsQqTq0nQjdaKQRECL1kbs192mP</credentials>
</user>
.
.
Replace the existing encrypted password with the new password.
Be sure to prefix the password with an exclamation point (!). For example:
<credentials>!mynewpassword123</credentials>
The password for the ias_admin
user should conform to following guidelines:
The minimum length is five alphanumeric characters.
At least one of the characters must be a number.
Passwords must be shorter than 30 characters.
Passwords can contain only alphanumeric characters from your database character set, the underscore (_), the dollar sign ($), and the number sign (#).
Passwords must begin with an alphabetic character. It cannot begin with a number, the underscore (_), the dollar sign ($), or the number sign (#).
See Also: "The ias_admin User and Restrictions on its Password" in the Oracle Application Server Installation Guide |
Start the Application Server Control.
After the restart, the Application Server Control will use your new Administrator (ias_admin
) password, which will be stored in encrypted format within the jazn-data.xml
file.
The performance metrics for a particular component show up as "Unavailable" in the Application Server Control Console.
Similarly, instead of a chart, one of the following messages (or a similar message) appears on the component Home page:
The underlying data for the CPU usage graph is not yet available
The underlying data for the Memory usage graph is not yet available
This problem often occurs immediately after the component is started.
Problem
Some metric data must be collected multiple times before the data can be displayed in the Application Server Control Console.
Solution
Verify that the component is up and running. If the component is down, restart the component.
If the component is up and running, wait at least five minutes to allow the necessary metrics to initialize, and then click the Refresh Data icon to refresh the data on the page.
From the Oracle Application Server Home page, you can quickly review the status of individual application server components, as well as the status of the overall application server instance itself.
Problem
Sometimes the Application Server Control Console indicates that the application server instance is down when components of the application server are up and running.
Solution
An Oracle Application Server instance is considered down when any one of its enabled components is down. For example, if one of your OC4J instances is down, the application server is considered down, even if the other components, such as Oracle HTTP Server, are up and running.
You can fix this problem by disabling components that are not in use. When a component is disabled, its status does not affect the status of the Application Server target. You can always enable the component at a later time.
To disable a component, click Enable/Disable Components on the Application Server Home page. Use the resulting page to determine which components you want to enable or disable for this application server instance.
See Also: "Disabling and Enabling Components" in the Enterprise Manager online help |
Before you can perform application server administrative tasks with Enterprise Manager, you must start Application Server Control.
Problem
When you attempt to start Application Server Control—or when the Enterprise Manager configuration assistant in the installation procedure attempts to start Application Server Control—an error occurs and the necessary processes are not started.
Solution
Problems starting Application Server Control are often caused by port conflicts on the host computer. In other words, a specific port number that the Application Server Control requires is in use by another application on the machine.
The first step in troubleshooting port conflicts is to identify which ports are in conflict. Some of the more common port conflicts that affect the Application Server Control include:
The port used in the Application Server Control Console URL
The port used by the Oracle Management Agent
The port used by OC4J Remove Method Invocation (RMI)
You can often identify a port conflict (or other startup problem) by reviewing the log files associated with these technologies or components. Table J-1 describes some of these component log files, which are installed in the sysman/log
directory of the Oracle Application Server home directory.
Table J-1 Log Files to Review When Troubleshooting Application Server Control Port Conflicts
Log File | Information To Look For in the Log File |
---|---|
|
Information and errors generated during startup of the Application Server Control |
|
Information and errors generated by OC4J RMI |
|
Information and errors generated by the Oracle Management Agent |
|
Additional information and errors generated by the OC4J instance used to deploy the Application Server Control |
After you identify a port conflict, you can modify the Application Server Control port number (if the port number can be reassigned).
From the Oracle Application Server Farm home page, you can view a list of the individual application server instances and OracleAS Cluster that are part of the Farm. To navigate to the Application Server Home home page for an instance, click the name of the application server target on the Farm home page. Similarly, the Cluster page provides a list of the application server instances that belong to the selected OracleAS Cluster.
Problem
When you click the name of the application server instance on the Farm home page, you receive one of the following errors in your Web browser:
In your Netscape browser:
There was no response. The server could be down or is not responding.
In your Internet Explorer browser:
The page cannot be displayed.
Solution
This error is most often displayed when the Application Server Control for the instance you selected is not running.
Note: In previous versions of Oracle Application Server, one Application Server Control was used to manage all the application server instances installed on a host. Starting with 10g (9.0.4), each application server instance, by default, requires an Application Server Control and Management Agent to be running from its Oracle home. |
To fix the problem, you must start the Application Server Control for the instance you want to manage:
Log on to the host where the application server instance resides.
Be sure to log in as the user who installed the Oracle Application Server instance.
Start the Application Server Control.
On UNIX systems, use the following command to start the Application Server Control:
ORACLE_HOME/bin/emctl start iasconsole
On Windows systems, use the Services control panel to start the Application Server Control service.
OracleAS Farm is a set of Oracle Application Server instances that share a common repository. If the instance you are managing is part of an OracleAS Farm, the Application Server Control Console URL displays the Farm page. The Farm page lists the application servers and OracleAS Cluster that are part of the Farm.
Problem
Sometimes the Application Server Control Console URL does not display the Farm page. Instead, Enterprise Manager displays the Application Server Home page. In the General section of the page, the Farm field indicates that the Farm is "Unavailable." Alternatively, the Farm field indicates that the "Infrastructure database is unavailable."
Solution
In most cases you can solve this problem by making sure that the OracleAS Metadata Repository database or Farm repository database is up and running. You can verify the status of the database by using one of two methods:
Display the Oracle Enterprise Manager Database Control Console for the Infrastructure Oracle home. The Database Control Console provides you with a Web-based user interface for monitoring and administering the Infrastructure database. From the Database Control Console, you can obtain the status of the database.
Use SQL*Plus to connect to the database and verify that it is up and running.
If the database is down, start the database and then navigate to the Application Server Control Console URL.
If the OracleAS Metadata Repository database is up and running, make sure the Oracle Internet Directory component of your OracleAS Identity Management installation is also available. Display the Application Server Control Console for the OracleAS Identity Management installation and check to be sure the Oracle Internet Directory component is up and running.
If both the OracleAS Metadata Repository database and Oracle Internet Directory are up and running, the problem may be related to port conflicts, which can affect the Oracle Process Manager and Notification Server (OPMN). Check the OPMN log files to identify the potential port conflict.
If you identify a port conflict, modify the port number (if the port number can be reassigned).
If your application server instance is part of an OracleAS Farm, some components of your application server instance may require access to the Identity Management components. Specifically, they may need access to Oracle Internet Directory.
Problem
When you log in to the Application Server Control Console, the following error message appears:
Unable to Connect to Directory Server:javax.naming.CommunicationException
Solution
This problem is caused when the Oracle Internet Directory component is down or unavailable. Verify that Oracle Internet Directory is up and running and start it if necessary.
For example, log in to the Identity Management host and enter the following command in the Infrastructure Oracle home to start the Oracle Internet Directory:
opmnctl startproc ias-component=OID
See Also: Oracle Process Manager and Notification Server Administrator's Guide for more information about starting and stopping OPMN components, such as Oracle Internet Directory |
Using Application Server Control, you can manage Oracle Application Server from a browser. As a result, you can manage your application server instances remotely as long as you have access to the network. In addition, multiple administrators can manage your application server instances.
Problem
In some cases, you may see the following error message displayed in your browser window:
The SMISession has been invalidated. Resolution: Please close the current SMISession, start another one and reapply the actions
Solution
To resolve this issue, click the Refresh Data icon located to the right of the time stamp, or close and reopen the browser to start a new session. This error can be the result of multiple users performing conflicting configuration actions on a single Enterprise Manager Application Server Control at the same time.
Oracle Application Server includes a version of the Oracle Management Agent that gathers monitoring data for the Application Server Control Console.
Problem
The Management Agent generates "out of memory" errors while collecting application server metrics.
Solution
Use the following procedure to increase the amount of memory available to the Management Agent Java Virtual Machine (JVM). The default value is 64 MB:
Use a text editor to open the following configuration file in the application server Oracle home:
(UNIX) $ORACLE_HOME/sysman/config/emd.properties (Windows) %ORACLE_HOME%\sysman\config\emd.properties
Locate the following entry in the emd.properties file:
agentJavaDefines=-Doracle.dms.refresh.wait.time=1000
Add the following qualifier to the agentJavaDefines
property to increase the available memory to 128 MB:
-Xmx128M
Restart the Application Server Control.
Application Server Control is the preferred management tool for most of your Oracle Application Server management tasks. However, you can still accomplish your management tasks using various command line tools.
Problem
If you use command-line tools to make administration or configuration changes to an Oracle Application Server instance (for example, if you use the dmctl
applyarchiveto
command), the changes are not reflected in the Application Server Control Console until after the Application Server Control cache is cleared.
Solution
To clear the cache, click the Refresh Data icon, which is located to the right of the time stamp, or close and reopen the browser to start a new session.
You can use the emctl secure iasconsole
command to configure the Application Server Control so it uses HTTPS secure communications.
However, after you configure security for the Application Server Control, you may get intermittent problems when using Microsoft Internet Explorer 6.0 or a later release.
Problem
Microsoft Internet Explorer has known issues with trying to reuse SSL connections after they have timed out. Due to this limitation, users connecting to Application Server Control using Internet Explorer, may see intermittent errors. Some examples of the errors include the following:
500 Internal Server Error
when deploying a J2EE application
Error: Processing already completed
after responding to a confirmation message
The graphics in the HTML version of Topology Viewers do not appear
Solution
To work around these SSL timeout errors, you can upgrade all browsers to use the correct Microsoft patches. For information about the Internet Explorer problem, its workarounds, and links to updates to Internet Explorer 6.0 and later, see the following:
Browser displays a message saying that the Session has expired
.
Problem
Opening multiple browser windows to view different Application Server Control Consoles on the same host may cause the browser to post a Session has expired
message if you switch between the browser windows. For example, you are viewing one Application Server Control Console located at:
http://mgmthost1.acme.com:1156/
You then open another browser to view an Application Server Control Console located at:
http://mgmthost1.acme.com:18100
As you switch between the two browser windows, you might receive a Session expired message
. This condition can occur with either Netscape Navigator or Internet Explorer.
Solution
To avoid this problem, start a new browser instance from the desktop and close any new windows opened from the original browser session. If you are using Netscape 7, you will need to create a new Netscape Profile for any additional browser windows.
You can specify whether you want to use the HTML Only version of the Topology Viewer or the Java applet version. To use the Java applet version, you must have the correct Java Plug-in support and proxy settings.
See Also: "Setting the Topology Viewer Preferences" in the Enterprise Manager online help |
Problem 1
The Java applet version of the Topology version requires Java Plug-in release 1.4 or later. Typically, your browser will prompt you to download the required version of Java Plug-in. However, in some browsers, you may not be prompted to download the plug-in, or you may be directed to an invalid URL. Without the correct plug-in support, the Topology Viewer applet does not load.
Solution 1
You must manually download and install the plug-in from http://java.sun.com/products/plugin/
.
Problem 2
If the Topology Viewer applet does not load, then look in the Java Plug-in Console for errors. Typically, the cause of these errors is the Java Plug-in could not access the proxy server.
You can start the Java Plug-in Console in one of three ways:
Windows system tray
Right-click the Java Console icon from the system tray, and select Open Console.
Microsoft Internet Explorer
From the Tools menu, select Sun Java Console.
Netscape
From the Tools menu, select Web Development > Java Console.
Solution 2
Configure the proxy settings to automatically detect settings and disable the use of the automatic configuration script. The details of this configuration varies from browser to browser.
See Also:
|
You can use Grid Control to discover Oracle Application Server instances.
Problem
If Grid Control release 10.1.0.2 or 10.1.0.3 is used to discover a release 9.0.4 or 10.1.2 Oracle Application Server instance, configuration changes, such as enabling or disabling components and creating or deleting OC4J instances in the Oracle Application Server instance will not be propagated to the central Management Agent and the Grid Control Console.
Solution
In order to propagate these changes to the Grid Control Console, remove the Oracle Application Server and its components from Grid Control with the following steps. When you remove the components from Grid Control, any collected data in the Management Repository will be deleted.
Click the Targets tab, and then All Targets subtab.
Select the Application Server instance, and click Remove.
Select the BC4J component for the Application Server instance, and click Remove.
Ensure that all targets have been deleted:
Click the Management System tab.
In the Management Services and Repository Overview page, in the General section, click the link Deleted Targets.
Once the targets are deleted, perform the following steps in the Application Server Control:
Navigate to the Application Server Home page for the Oracle Application Server instance.
From the Application Server Home page, click the Infrastructure tab.
In the Grid Control Management section, click Configure.
In the Configure Grid Control page, select the appropriate Management Service, and then click OK.
The Oracle Application Server and its components will now appear in the Grid Control Console. You can now enable or disable components and create or delete OC4J instances in the Oracle Application Server.
When you use Application Server Control Console to monitor targets, such as an instance of OracleAS Portal, running in an environment configured for SSL, some performance metrics may not display.
To correct this problem you must allow the Application Server Control to recognize the Certificate Authority that was used by the Web site you are monitoring to support HTTPS. You must add the Certificate of that Certificate Authority to the list of Certificate Authorities recognized by the Application Server Control.
Problem
If you are searching the Log Repository for log entries that occurred during a specific time frame, you might have problems displaying the pop-up date selection window.
Solution
This problem can be caused by customizations that you made to Microsoft Internet Explorer browser. For example, if you installed third-party pop-up blocking software, the browser may not be able to display the date selection window.
To work around this problem, deinstall or disable the browser customizations. Alternatively, enter the date directly into the date field, using the following date format: MM/DD/YY
.
See Also: "Searching the Log Repository" in the Application Server Control Console online help |
The following sections describe problems and issues when using Application Server Control to manage an OC4J instance and the J2EE applications you deploy:
Redeploying WAR Applications with Application Server Control
Deployment Performance in Internet Explorer and Netscape Navigator 7.0
You use the OC4J Security page in the Application Server Control Console to configure various security settings for your deployed J2EE applications.
Problem 1
After making changes on the OC4J Security page, the changes do not seem to affect the deployed application.
Solution 1
After you make changes on the OC4J Security page, you must restart the OC4J instance in order for the changes to take effect. For example, if you add the user admin
user and the administrators
group as described in Section J.3.2.2, you must restart the OC4J instance to complete the procedure.
Problem 2
OC4J security employs a user manager to authenticate and authorize users and groups that attempt to access a J2EE application. One of the user managers that can be used to designate the users and groups for an application is the JAZN user manager.
With Application Server Control, you can specify that the JAZN user manager be associated with an application. Using the Deploy Application: User Manager page of the Application Server Control Console, you can specify that the application use either a JAZN XML configuration or a JAZN LDAP configuration.
When you use Application Server Control Console to specify an XML-based JAZN configuration, the following line is entered into the orion-application.xml
file:
<jazn provider="XML" location="./jazn-data.xml" />
When you use Application Server Control Console to specify an LDAP-based JAZN configuration, the following line is entered in the orion-application.xml file:
<jazn provider="LDAP" default-realm="sample_subrealm" />
Some applications may prefer to specify a JAZN configuration by providing a path to a jazn.xml
file, but Enterprise Manager does NOT support this type of JAZN configuration. This type of JAZN configuration would be specified as follows in the orion-application.xml file:
<jazn config="jazn.xml"/>
If you manually specify this type of JAZN configuration in the orion-application.xml
file, you will either be unable to use the Application Server Control Console OC4J Security page or you will experience problems even after apparently using the page successfully.
Solution 2
Do not manually configure JAZN by providing a path to the jazn.xml
file.
For more information about user managers specifying users and groups for a J2EE application, see Oracle Application Server Containers for J2EE Security Guide.
From the Application Server Control Console, you can deploy J2EE applications to Oracle Application Server Containers for J2EE (OC4J).
Problem
When you are attempting to deploy an OC4J application using the Application Server Control Console, you may receive the following error:
Deployment failed: Nested exception Root Cause: Lookup error: javax.naming.NoPermissionException: Not allowed to look up java:comp/ServerAdministrator, check the namespace-access tag setting in orion-application.xml for details;
Solution
This error may appear if the user manager for the OC4J default application does not include the user admin
and the group administrators
.
To view or define the users and groups for the default application user manager:
Navigate to the OC4J home page for the OC4J instance you used to deploy your application.
Click Applications to display the list of application deployed in the selected OC4J instance.
Click the Default Application Name, which appears at the top of the Applications page.
Enterprise Manager displays the OC4J Application home page for the default application.
Scroll to the bottom of the page and click Security.
Enterprise Manager displays the Security page, which lists the Groups and Users.
On the OC4J Applications Page in the Application Server Control Console, you can deploy EAR files (applications with a file type of .ear
) and deploy WAR files (web applications with a file type of .war
).
To deploy a WAR file using the Application Server Control Console, click Deploy War file on the OC4J Applications Page.
The first time you deploy a WAR file, Enterprise Manager launches a deployment tool that automatically wraps the WAR application into a J2EE application (.ear
file) before deploying it. The .ear
file that Enterprise Manager creates to deploy your WAR file contains an application.xml
file that describes the application modules. The .ear
file is given an application name that you supply when you step through the deployment tool. After the WAR application is deployed, the name of the new application (.ear
file) appears in the Deployed Applications table.
Problem
After you have deployed a WAR file using the Application Server Control Console, it cannot be redeployed by selecting the application (.ear
file) on the OC4J Applications Page and clicking Redeploy.
Solution
To redeploy a WAR file using Application Server Control Console, you must undeploy the application first, then deploy it again by following these steps:
In the Deployed Applications table on the OC4J Applications Page, select the application (.ear
file) in which the WAR file was wrapped and deployed.
Click Undeploy.
Click Deploy War file. In the deployment tool, specify the same application name as you specified the first time for the application (.ear
file) in which the WAR file was wrapped and deployed.
After the WAR application is deployed, the name of the web application (.ear
file) appears in the Deployed Applications table.
Problem
If you attempt to deploy an OC4J application while using Microsoft Internet Explorer or Netscape 7.0, the file upload may take an extremely long time (for example, 10 minutes for a 45 MB .ear
file as compared to 15 seconds with Netscape 7.1).
Solution
If you are using Netscape Navigator, upgrade to Netscape 7.1.
If you are using Internet Explorer, refer to the following Microsoft knowledge base article, which addresses this problem:
http://support.microsoft.com/default.aspx?scid=kb;en-us;329781
Problem
When attempting to deploy a large application (greater than 50 MB EAR or WAR file), the default configuration of the Distributed Configuration Management (DCM) Daemon is insufficient. Attempting to deploy such a large application results in an "Out Of Memory" error.
Solution
Use the Application Server Control Console to increase the memory for the DCM Daemon component:
Navigate to the Application Server home page in the Application Server Control Console.
Locate the java-parameters
<data>
tag in the DCM Daemon configuration section of the file:
<ias-component id="dcm-daemon" status="enabled" id-matching="true"> <process-type id="dcm-daemon" module-id="DCMDaemon"> <process-set id="dcm" numprocs="1"> <module-data> <category id="start-parameters"> <data id="java-parameters" value="-Xmx256m -Xrs -Doracle.ias.sysmgmt.logging.loglevel=ERROR -Djava.net.preferIPv4Stack=true -Djava.io.tmpdir="$TMP""/> . . . </ias-component>
Replace the string -Xmx256m
with the string -Xmx512m
.
For example:
<data id="java-parameters" value="-Xmx512m
This new value increases the memory assigned to the DCM Daemon from 256 MB to 512 MB.
Click Apply to save your changes.
Open a terminal window (UNIX) or a DOS Command window (Windows) use the following commands to reload the OPMN configuration file, restart DCM, and restart the Application Server Control.
On UNIX systems:
ORACLE_HOME/opmn/bin/opmnctl reload ORACLE_HOME/opmn/bin/opmnctl restart ias-component="dcm-daemon" ORACLE_HOME/bin/emctl restart iasconsole
On Windows systems:
ORACLE_HOME\opmn\bin\opmnctl reload ORACLE_HOME\opmn\bin\opmnctl restart ias-component="dcm-daemon" ORACLE_HOME\bin\emctl restart iasconsole
Try deploying the application again.
Problem
Depending upon the size and number of applications you deploy to your OC4J instance, you might experience "out of memory" errors.
Solution
Adjust the Java Virtual Machine (JVM) heap size for your OC4J processes.
See Also:
|
You can find more solutions on OracleMetaLink, http://metalink.oracle.com
. If you do not find a solution for your problem, log a service request.
See Also: Oracle Application Server Release Notes, available on the Oracle Technology Network: |