Configuration and Administration

About Scalability

  • Minimize the performance overhead of cache synchronization between application servers in the same cluster.
  • Provide load balancing among application servers in the same cluster.

After setup, view functional and domain distributions on the Scalability Overview page.

Application Functions

Note: This is for on-premise implementations only; it is not valid for Cloud implementations.

Oracle Transportation Management currently supports the following applications functions which can be accessed via Configuration and Administration > Cluster Management > Application Functions. You must have DBA rights to access this.
  • Integration: applies to any data being received into Oracle Transportation Management electronically. A cluster for Integration can contend with user actions on other clusters, so creating an integration cluster is not always recommended.
  • Process Control: each process control task is associated with an Application Function. Clusters can be dedicated to processing one or more scheduled or recurring process control tasks.

Application Servers

You define a Machine URL for a specific Application Server ID via Configuration and Administration > Cluster Management > Application Servers. You must have DBA access to access this.

Cluster Manager

The Cluster Manager is where you configure the settings for your cluster of application servers.

Certain Oracle Transportation Management property files must be set for the application server scalability pages to be enabled in Oracle Transportation Management.

Message Flow

In addition, Oracle Transportation Management needs to ensure that whatever data resides in memory is synchronized across clusters. In Oracle Transportation Management, JMS is being used for this purpose.

Since there can be multiple application servers in a cluster, the information between the servers needs to be coordinated with the database. This is accomplished with SQLNet, which guarantees data integrity through data locking.

Logging Support for Clustered Application Servers

If you have enabled application server clustering, you can indicate which system logs you want to analyze. Choose Configuration and Administration > Process Manager > Logs > System. The Machine list box enables you to indicate which server to search or you can search all servers. This is particularly important when you are load balancing between servers since you may not be sure which machine processed your request. This field is only available once scalability has been enabled through the Oracle Transportation Management properties file.

If you have enabled clustering, you can also monitor the performance of clusters through the Performance log. Choose Configuration and Administration > Process Manager > Logs > Performance. You can specify the Scalability Mode you want to monitor for a specific domain:

  • Current: shows performance statistics for the current machine.
  • Machine: shows performance statistics for a specific machine on the Oracle Transportation Management application server network. The corresponding Machine field lists the servers on the network from which to choose.
  • Cluster: show performance statistics for one or more servers that are defined in the cluster. Specify a domain.
  • Network: shows performance statistics for all servers on the Oracle Transportation Management application server network.

Failover Support

If for some reason one cluster fails, all processes from the Web server can be assigned to another cluster. This switch is seamless to any Oracle Transportation Management user.

  • If all application servers in a cluster fail, the user receives an error.
  • When an application server is brought back up, it registers itself with all the other servers to re-enable load balancing across clusters.
  • You must have a minimum of two servers in a cluster for this type of failover support to occur.

One application server can be restarted without having to restart all servers in a cluster.

Note: Milestone monitor functionality pre-dates scalability architecture and therefore is not designed to handle failover.

Note: This type of failover support has been deprecated as of the 6.4.3 release and will be removed in a later release.

Related Topics