Groups

Groups are objects to further group the related tasks. A task may belong to zero or more groups. If the task belongs to one or more groups then one of the groups should be it's Primary Group. The primary group is mainly because for the visualization purposes. Other than that, all groups are the same for that task. Groups are useful for managing the parallelization. For example;

Suppose you have 4 different systems and you want to extract data from those systems and load to the target system every day. Which means you have four source systems and a target system. You both have to control the parallel sessions on the individual source systems and the parallelization on the target system. You can design the tasks like belonging to the extraction group of individual system as well as belonging to a more general parent extraction group. That way you can both control the source system and target system resources.

The parallazition logic by default considers the minimum available slot logic. This means if a task belongs to Group1 and Group2 then the available slot is calculated like min(Group1.parallel, Group2.parallel) This behavior is default but there is an issue for enhancing this logic.