1.1 Conceptual Overview

With constant advances in the capability, scalability, and affordability of computing and communications technology, a few noticeable trends emerge in the way that corporations manage their operations:

  • The number of computers in the workplace that are used to accomplish day-to-day tasks is greatly increasing.

  • The diversity of computers in the workplace is increasing and now includes desktop computers, laptops, servers, and mobile devices.

  • More organizations are opening branch offices in remote locations. Those branch offices still require access to the data and computing resources of the central office.

  • More organizations are using data center services that specialized companies provide.

As a result of these trends, managing a company's IT infrastructure is both complicated and mission-critical. An administrator has to monitor computers and software, collect and analyze performance data, and carry out actions while rarely having direct physical access to the computers themselves. For example, an IT administrator simultaneously might have to manage the power consumption of servers in a data center, the security settings for the operating systems that run on office desktops, and the configuration options for applications that the employees use to get work done.

Another major complicating factor is the diversity of the computers to be managed. This diversity manifests itself in several ways, including the following:

  • Different categories of computers: desktop computers, laptops, servers, mobile phones, and more.

  • Different processors, for example, 32-bit or 64-bit chips.

  • Different operating systems and applications that are installed on these computers.

To simplify the management of a heterogeneous and widely-distributed computers, it is necessary to provide both a common mechanism for retrieving and manipulating data and a common format for representing that data. In this way, a single management application or interface can be used across the entire organization. That is, the IT administrator knows how to work with all of the computers, and any new computer that is added to the organization is compatible with the existing management applications and tools. All device manufacturers have to share this consistent representation for it to be effective, so it can support a variety of devices with different capabilities.

There are two different approaches for Windows Management (WM):

  • Object/resource-based management

  • Command/task-based management

Object-based management uses class objects and instances to represent data. The CIM Infrastructure Specification ([DMTF-DSP0004]), provides consistent data representation. CIM is a conceptual model that is not bound to any particular implementation. It also allows for vendor extensions, so any system that exposes CIM-compliant data can be accessed in a consistent manner. If vendors require additional functionality, they can extend the standard CIM schema.

WM protocols enable management applications and tools to access CIM data remotely. Section 2.1 provides more details about how WM protocols retrieve CIM data.

Windows Management protocols enable management applications and tools to access CIM data remotely. The data can be retrieved through one of the three independent, remote WM protocols:

  • The Windows Management Instrumentation Remote Protocol, as specified in [MS-WMI].

  • The Web Services Management Protocol Extensions for Windows Server 2003 operating system, as specified in [MS-WSMAN].

  • The Web Services Management Protocol Extensions for Windows Vista operating system, as specified in [MS-WSMV].

These three remote protocols enable network communication between the management application and the CIMOM. For more detailed information, see section 2.1.

A different management approach is the command/task-based management, which uses commands and command shells to perform a particular administration task. These commands process input streams and are framed into pipelines that are similar to UNIX pipelines, with the difference that these pipelines represent commands and parameters in an abstract structured way, independent of any higher-layer syntax or semantics, by using an XML representation.

The command execution engine on the server executes a task specified by a set of commands that are run through a shell s. The results/outputs are relayed back to the client as streams that are then sent to higher layers.

The following Windows Management protocols support this type of management:

  • PowerShell Remoting Protocol (PSRP), as specified in [MS-PSRP].

  • PowerShell Remote Debugging Protocol, as specified in [MS-PSRDP].

  • The Web Services Management Protocol Extensions for Windows Vista, as specified in [MS-WSMV].

The pipelines can be executed either by the REST-based approach or the session-based shell approach, as described in section 2.1.