A push removal step references a step definition run to determine what objects should be removed from the target collection on the Agentry Client. A removal step can be executed either once, or iteratively based on the number of users logged in to receive push data. A removal step is run as a part of the back end polling performed by the push. The step referenced by a push removal step definition is expected to return the key property of any object instance to be deleted from the Client. If defined to run once per poll period it is also expected to return the Client user ID from which the object will be removed.
A removal step may be run once per poll of the back end system, or once per user per poll period, depending on how it is defined. When a removal step is run once per user, data returned by the step will be organized according to each user. Note that this scenario can result in a large number of executions of the step per poll, as the number of users logged in into the Server is commonly in the hundreds or more in a production environment. When run in this manner, the user ID value for each user is accessible via the <<userID>> SDML data tag.
When a removal step is defined to run one time, it will be executed once per poll period, regardless of the number of users connected to the Server. In this situation, the data returned by the removal step should also include the user ID as entered on the Client, indicating which client will receive the key property for the object to be removed. It is recommended that this is how most, if not all removal steps are defined within a push as it is a more efficient model of data synchronization. When run in this manner individual user ID’s are not available to the step.