Enumerates the options for how running replication agent threads are terminated when an unanticipated shutdown occurs.
Assembly: Microsoft.SqlServer.Replication (in Microsoft.SqlServer.Replication.dll)
An unanticipated shutdown can occur because of a system runtime failure or a user initiated abort request.
The policy should be chosen depending on application state requirements. An asynchronous policy may be more responsive during a shutdown but can require the use of other techniques to return the runtime environment to an expected state. A synchronous policy can help avoid inadvertent interruption of updates to these critical global runtime states when a replication agent is invoked synchronously from an application.