Use this instruction to mark the point at which saved data
values replace the current values in the data fields that are designated
for rollback. RollbackCheckpoint configuration
options determine both which data fields are to be rolled back and
where work item processing resumes after the rollback.
Here are some guidelines for deciding which data fields
to roll back:
- Avoid rolling back values you can use to help you determine whether
an error occurred (or the nature of the error that occurred).
- Roll back field values that affect subsequent processing.
To specify a RollbackCheckpoint system
function:
- From the CheckPoint Palette, drag a RollbackCheckpoint
step onto the workflow map after the step where the values were
saved with BeginCheckpoint.
- Under Resume Processing Expression,
enter a Boolean expression that determines where processing resumes
after the rollback occurs. Step resumption occurs in the
following way:
True expression
|
The step that immediately follows BeginCheckpoint runs.
This step runs again even if that step is this RollbackCheckpoint.
This option is the default option.
|
False expression
|
The step that immediately follows RollbackCheckpoint runs.
|
- To do some additional processing before the workflow returns
to the checkpoint, select a map that contains the process.
- Indicate the rollback status for each user-defined data
field. By default, all user-defined data fields are listed in
the Roll Back Fields column. Use the arrow
keys to move the fields between panes as needed. As
shown in the following table, the rollback status determines what
happens to the current value in each data field when this RollbackCheckpoint system
function runs:
Data field is in Roll Back Fields column
|
The current value is replaced with the value
that was saved for the data field when BeginCheckpoint ran.
|
Data field is in Non Roll Back Fields column
|
The current value is retained.
|