Plug-ins

IBM® UrbanCode Deploy plug-ins provide tools for creating component processes.

Plug-ins consist of configurable steps, which can be thought of as distinct pieces of automation. By combining steps in the IBM UrbanCode Deploy editor, you can create fully automated deployment processes. In addition to basic plug-ins, others integrate many third-party tools into IBM UrbanCode Deploy, such as application servers and software configuration management products. For example, the Tomcat and WebSphere® plug-ins provide steps that start and stop those servers, install and uninstall applications, and run other tool-specific tasks. Finally, you can write your own plug-in (see Creating plug-ins).

A plug-in consists of a number of steps. These steps vary from plug-in to plug-in. Each step consists of a number of properties, a command that runs the step, and post-processing instructions (typically used to ensure that expected results occurred). Step properties can serve a wide variety of purposes, from providing information to the command, to supplying some or all of the actual command itself. When you create a process, you drag steps onto the editor's design area and define the properties as you go. Property values can be supplied when you define a component process or at runtime. The process flow is defined by drawing connections between steps. In the following illustration, you can see a series of plug-in steps and the connections between them. For information about creating component processes, see Component processes; for information about creating your own post-processing scripts, see Post-processing scripts.

For example, the following figure shows a process that uses the IBM WebSphere Application Server plug-in to deploy an application. It includes the generic Download Artifacts step that retrieves the component files from CodeStation. Then, it uses several steps from the plug-in to stop, update, and start the application.

A simple component process that uses the WebSphere Application Server plug-in to deploy an application

At deployment time, component processes are run by agents that are installed in the target environment. For a process to run successfully, the agent must have access to all resources, tools, and files that are required by the plug-in steps in the process. When you install an agent, ensure that:

For information about installing agents, see Installing agents from the command line.


Feedback