Managing stateful operations


You use the transaction API to manage stateful configuration API operations. Before you can perform any stateful operation in system configuration APIs, you must initiate a transaction via the transaction API.

For configuration changes to take effect, you must commit your operations. When you commit a transaction, the system creates a copy of the current configuration settings to avoid conflict with changes made by other users.

You can roll back a change that you have committed until you terminate the transaction. After the transaction is terminated, either by you, automatically, or because the connection request times out, you cannot roll back any configuration changes that you made. To prevent the transaction from timing out, use the corresponding command.

Identifiers for stateful transactions

The system assigns every initiated transaction a transaction identifier (tid), which you must use in subsequent transaction operations.

A transaction identifier is returned in the response of any successful request that you make to initiate a transaction on the system. The transaction identifier must be specified in subsequent requests to the configuration APIs.

Workflow for performing stateful operations

Use the following procedure as a model for performing stateful operations:

  1. Initiate a transaction and receive a transaction identifier in the response.
  2. Perform any operations to configure features, specifying the transaction identifier in each request.
  3. Commit the transaction.
  4. Terminate the transaction.
     If you must revert to the previous configuration settings, roll back the transaction.

Related topics

Stateful-versus-stateless-operations-in-configuration-APIs
Composing-requests-in-configuration-APIs

 

Tip: For faster searching, add an asterisk to the end of your partial query. Example: cert*