This chapter describes issues associated with Oracle User Messaging Service (UMS).
It includes the following topics:
This section describes general issue and workarounds. It includes the following topics:
Section 6.1.1, "Extension Driver has Different Target Types for 11g and 12c"
Section 6.1.4, "User Messaging Service URLs Unavailable After Restart"
Section 6.1.6, "Oracle Service Bus Applications Display Health State as Warning"
Section 6.1.7, "User Messaging Service Upgrade from 11g to 12c may Fail"
Section 6.1.8, "Duplicate Messages Rendered in Twitter Driver"
Section 6.1.9, "Exceptions Occur while Registering Multiple Drivers and Access Points"
In Oracle Enterprise Manager Fusion Middleware Control 12c, the User Messaging Service Extension driver will be displayed under the User Messaging Service folder, instead of the Application Deployments folder in the left navigation pane. The driver performance data is also available for the extension driver. This is an expected behavior.
In a cluster environment, the node manager may fail to start if you have configured Oracle User Messaging Service (UMS) and Oracle HTTP Server in a domain with Oracle Real Application Clusters (RAC) multi data sources.
UMS uses about 100 connections per data source if all UMS drivers are deployed and running. Therefore, in a clustered environment with RAC setup, you may have to increase the maximum number of connections allowed on the database server. Set this value to the sum of maximum number of connections per data source for each WebLogic Server. For example, when Oracle RAC is used with three nodes, that is two WebLogic Servers with three Oracle RAC data sources, set the maximum number of connections to 600 (2 x 3 x 100).
When no metric data is found (for example when no messages have been sent or received after server setup), the Metrics Performance page will display Unavailable. This is not a problem with the software, and the Performance reporting is operating properly. As soon as Send and Receive traffic exists, the Performance page will display results normally.
Upon restarting the User Messaging Service server (usermessagingserver) from Oracle Enterprise Manager Fusion Middleware Control or through Oracle WebLogic Console, you may get an error: Error 503--Service Unavailable when attempting to access any URLs served by the User Messaging Service server, such as the User Preferences UI (/sdpmessaging/userprefs-ui) or the various Web Services endpoints. This error occurs intermittently in cases when the Oracle WebLogic Server is heavily loaded (such as with a SOA instance). To work around this issue:
Restart the User Messaging Service server again (two or more restarts may be required).
If multiple User Messaging Service server restarts are not sufficient, then restart the entire Oracle WebLogic Server instance.
In a cluster environment, the JNDI lookup fails for remote UMSJMSServer.
When UMS is deployed in a cluster with two or more nodes, the JMS resource called UMSJMSServer must be targeted to each managed server in that cluster. A scenario where two or more UMSJMSServer resources are targeted to the same managed server is not supported.
The health state of the following Oracle Service Bus (OSB) applications are displayed as 'Warning' in both the Admin Server Console under deployments and in the Admin Server logs:
Service Bus Email Transport Provider
Service Bus File Transport Provider
Service Bus FTP Transport Provider
Service Bus SFTP Transport Provider
In some scenarios, while upgrading UMS from 11g to 12c, the upgrade may fail with the following error:
[2014-04-10T20:38:14.915-07:00] [UCSUMS] [ERROR] [] [upgrade.UCSUMS.UCSUMS_CONFIGURATION_PLUGIN] [tid: 70] [ecid:435559f9-7615-48f8-8e80-950a7f10e152-00000001,0] [[com.jcraft.jsch.JSchException: verify: false
The com.jcraft.jsch.JSchException: verify: false exception is intermittent. As a workaround, try to upgrade again.
Each time the Twitter driver is restarted, the last 20 tweets are fetched for the configured user. This means, that the first receive attempt after restarting the application might return some messages that have already been seen previously. As a workaround, if you want to handle every tweet only once, you can use the header containing the Twitter Id.
When multiple drivers are trying to register at the same time, the registration fails due to the following Internal Exception:
java.sql.SQLIntegrityConstraintViolationException: ORA-00001: unique constraint (SYZSLC07PAT_UMS.QUEUE_PK) violated
As a possible workaround, stop and start the email driver through the Oracle Enterprise Management Console. This triggers a new registration of the driver.
When multiple clients are trying to register access points at the same time, the registration fails due to the following Internal Exception:
java.sql.SQLIntegrityConstraintViolationException: ORA-00001: unique constraint (SYZSLC07PAT_UMS.ACCESS_POINT_PK) violated
As a possible workaround, stop and start the application that triggered the flow that resulted in this exception. For example, restart the application that triggers the SOA UMS Adapter to register the AccessPoint. Please note that it is not the SOA UMS Adapter itself that needs to be restarted but the application hooking up to the SOA UMS Adapter that must be restarted. This triggers the SOA UMS Adapter to register the AccessPoint with UMS.
While upgrading the UMS schema using Oracle Fusion Middleware Upgrade Assistant, you will notice that - the connect button that is used to obtain connection to the database, and the drop-down list that is used to populate the available UMS schema names - are missing. The workaround is to manually enter the value for schema name.
This section describes configuration issues and their workarounds. It includes the following topics:
Before configuring any User Messaging Service Driver (such as the Email Driver), to connect to a remote gateway using SSL, ensure that the SSL Trust Store is properly configured as described in "Configure Keystores" in Oracle WebLogic Server Administration Console Online Help.
Ensure that the value of the JVM system property (javax.net.ssl.trustStore) set in $DOMAIN_HOME/bin/setDomainEnv.sh (or Windows equivalent file) points to the correct trust store that you want to use.The Java Standard Trust Store is located at:
$JAVA_HOME/jre/lib/security/cacerts or $BEA_JAVA_HOME/jre/lib/security/cacerts
With the default out-of-the-box configuration of SSL trust store, that is with the Java Standard Trust Store, the UMS driver will be able to connect to the Oracle Beehive Email Server over SSL. Note that in some installations, for example, when you have SOA installed, the Java Standard Trust Store is replaced by a Demo Trust Store. In such situations, the Trust Store may not contain the valid root certificate needed by Oracle Beehive Email Server. To resolve this issue, follow the instructions for using the correct SSL trust store. Replacing the DemoTrust keystore in the setDomainEnv.sh file (or Windows equivalent file) with the Java Standard SSL trust store will enable UMS email driver to connect successfully over SSL to the Oracle Beehive Email Server.