Glossary

Component model:

An XML representation of a component. The component model is a bill of materials of the applications that includes:

• A list of resources used by the application

• Install steps

• Uninstall steps

• Dependencies

Component:

A component is a unit that is deployed to hosts in the provisioning software. It comprises a build between a component model and a specific set of resources for an application, such as a custom Java application or an infrastructure software package. A component can comprise multiple versions of an application. In this case, each checked-in version of the application is represented by a separate component build.

Deployment:

The execution of a plan, which automatically distributes a component or components included in the plan. The deployment follows installation instructions set in the plan. Deployment is always preceded by a preflight. If the preflight is successful, the Plan Executor will actually deploy/install components to the target host.

Comparisons:

A provisioning software feature that searches for and identifies differences between hosts. There are three types of comparisons: Model to Install, Model-to-Model, Directory-to-Directory.

The install-to-install comparison examines the contents of two hosts' file systems and reports any differences.

The model-to-install comparison compares what the Master Server reports is installed on a host against what is actually installed on the host, and reports any differences.

A model-to-model comparison examines the deployment repository and history (stored on the Master Server) for two hosts, and reports any differences.

execNative:

An optional call out to custom scripts from a plan XML.

Gold server:

A reference (outside) host that contains files, directories, and other resources comprising an application you want to deploy. These resources are checked in (uploaded) to the provisioning software from the gold server.

Host:

A server connected to the provisioning software. A host can represent an individual server or a set of servers.

Host set:

A user-defined, logical grouping of hosts sharing one or more common attributes, such as physical location or functional group. A defined host set allows quick and easy application updates to all hosts in the set, as well as install-to-install comparisons between the hosts. Host sets can be nested.

Host type:

A base class of servers bound by a set of common attributes, all of which are user-defined. Hosts belonging to a host type set use particular attribute values in variable substitutions in all deployments. Host types can be used to categorize hosts into logical groupings, such as sets of applications required for the hosts, geographic locations of the hosts, or functional groups. Host types also facilitate host searches by allowing users to search for all hosts of a certain type.

Host search:

A query run on the repository that yields a list of hosts whose attributes match those specified in the query. You can use host searches to create lists of hosts that have the same host type, are running the same applications, are configured with the same subnet masks, and so on.

Local Distributor, or LD

An engine, installed on a machine with the provisioning software, that acts as a link between other objects, such as the Master Server to Remote Agents, the Master Server to other LDs, and another LD to Remote Agents. Designed to maximize bandwidth efficiency and speed, LDs can also provide secure network connections for navigating restricted environments.

Master Server, or MS:

The main processing engine in the provisioning software. It is installed on a dedicated machine at a central location, where it can connect to any of the data center environments. The MS provides the centralized data storage, data processing, and user interfaces.

Plan:

An XML representation of the deployment script that references components to be deployed. A plans include:

• Install/uninstall steps.

• Start/Stop components (optional).

• Verification steps (optional)

• ExecNatives (optional)

Plan executor:

The provisioning software's deployment engine. It handles preflights and actual deployments.

Preflight:

A test “deployment” of a plan to a simulated UNIX environment that:

• Checks availability and connectivity of target hosts.

• Confirms permissions.

• Validates dependencies.

A preflight finds and reports any errors or potential errors that could affect the actual deployment. Preflights always automatically proceed a deployment; they can also be run as standalone procedures that are not automatically followed by a deployment.

Remote Agent, or RA:

An engine installed on any machine in the provisioning software to which components are deployed. On that machine, the RA manages tasks such as installing software, controlling services, collecting information to send back to the MS.

Resources:

Files that are deployed to hosts when a plan is executed.

Resource types:

Built-in resource categories identified and supported by the provisioning software. When a resource is checked in, it must be tagged with the appropriate resource type. Resource types include files, directories, IIS Types, and COM+ components.

Snapshot:

A capture of resources stored on a Remote Agent during a deployment. The snapshot is used during comparisons between hosts (install-to-install) or between a host and the model of the host on the Master Server (model-to-install).

Verification step:

An optional URL or process test that can be run at the end of a deployment to ensure successful completion of the plan.