When a data store executes, it loads new data from its source and places this data into its defined fields.
Knowing when a Data Store loads new data can help you to determine the data that you are working with and when to schedule data view executions, or it can indicate the execution of an analysis.
Rollback
Rolling back a data store restores it to the state it was in prior to its most recent run, and deletes any data that was added to the data store during the most recent run.
Data view execution
Run
When a data view runs, it takes data from the data stores it has been pointed to and loads the data into its fields, based on the data view's data load configurations.
Rebuild
A data view rebuild deletes all data presently residing in the data view and then loads all data from its data stores, overriding the data view's data load configurations.
Knowing when a data view executes can help you to determine precisely what data it will allow you to explore. Knowing when to perform or schedule data view executions depends on the frequency at which the data view's data stores receive new data and also the frequency at which the data view is modified.
Rollback
Rolling back a data view will restore it to the state it was in prior to its most recent run. This will delete any data that was added to the data view during the most recent run.
Analysis execution
Run
When an analysis runs, it performs the enhancements and/or analytics defined in its design and then loads its data store outputs with new data. The data pulled into the analysis will be as specified by the analysis' inputs' data load configurations.
Rebuild
An analysis rebuild performs the same activity as a run, enhancing and/or analyzing data and then loading data store outputs. However, the difference is that all data from all inputs will be pulled into the analysis regardless of data load configurations.
Knowing when an analysis executes can help you to determine exactly what data is contained in a data store that the analysis is pushing to and, subsequently, what data is contained in a data view that data store loads. Determining when to perform or schedule analysis execution is primarily dependent upon the frequency at which its data store input receives new data.
Rollback
Rolling back an analysis will restore it to the state that it was in prior to its most recent run. This will also delete any data that was added to the analysis' data store outputs during the most recent run.
Process model execution
A process model can be run or rebuilt to carry out the tasks specified in its design, however, in general, you only need to save a process models. In most cases, process model execution will be triggered by the execution of another data stage.
For more information, see Process Models.
Run
When a process model runs, it performs the tasks defined in its design. If execute tasks are performed, they will run the specified data stages.
Rebuild
A process model rebuild performs the same activity as a run, however, if execute tasks are performed, they will rebuild the specified data stages.
Rollback
Rolling back a process model will restore it to the state that it was in prior to its most recent run. This will also rollback any executables that were triggered by execute stages tasks and as a result will remove any data that was loaded to data stages during the rolled back run.
Rollback child processes
When rolling back a process model, you have the option to Rollback Child Processes. When Rollback Child Processes is selected, the rollback will additionally rollback any process models that were executed by the current process model using an execute process task. This rollback of child processes is applied recursively, meaning that if the child processes have child processes, those processes will be rolled back as well.