共用方式為


About Control Codes

The Failover Cluster API uses control codes to define cluster operations and cluster event notifications. The API has a large number of built in control codes, and also supports custom controls codes. A Control code is a 32-bit integer that defines a cluster operation or an event notification. The primary way to perform operations on clusters and respond to cluster events, is to pass a control code to a control code function. A control code function initiates an operation on a cluster or notifies a resource DLL of a cluster event.

There are two types of control codes:

  • External control codes, which define control code operations. Applications and resource DLLs can initiate cluster operations by using external control codes.
  • Internal control codes, which define cluster event notifications. The Cluster Service uses control codes to notify resource DLLs about cluster events that affect specific resources and resource types. A resource DLL can respond to those cluster events by supporting the control code of the event.

In this section

External Control Codes

Details about external control codes.

Internal Control Codes

Details about internal control codes.

Control Code Architecture

Describes the structure of control codes. You need this information in order to create custom control codes.

The Failover Cluster API

Using Control Codes

Control Code Reference