共用方式為


Failover Cluster Control Codes

The Cluster API defines a small set of control code functions—one for each cluster object—as "generic" interfaces to a wide range of cluster operations. Control codes—32 bit values used as function parameters—determine the kind of operation a control code function will perform. Together, the control code functions and the control codes support nearly every kind of cluster functionality.

Control codes are 32-bit integers used to describe an operation on a cluster object. Control codes are categorized as internal or external:

It is important to become familiar with the capabilities of the control codes. Applications and resource DLLs should always use control codes, if possible, to perform operations on cluster objects. The control codes are designed to maximize security, protect the cluster database, and ensure resource DLL notification.

In this section

Failover Cluster Reference