Tag
A tag is an informative label attached to a Task. Tags carry no effect on planning or execution; they exist for reporting, filtering and aggregation.
A task may carry several tags. Tags are typically used to record which consumers depend on a task's output.
Attributes
| Attribute | Description |
|---|---|
| Name | Name of the tag. |
| Description | Free text. |
Typical use
The usual pattern is to tag tasks by the department or system that consumes their data. A shared task carries every tag that applies to it:
ETL_CUSTOMER— tagged Sales and AccountingETL_ORDER— tagged SalesETL_PROFITLOSS— tagged Accounting
Downstream tasks can then be tagged according to the actual use of the data rather than according to where it happens to be computed.
What tags are used for
- Filtering the Monitor. A tag filter restricts the task list to the tasks feeding one consumer. This is the quickest way to answer whether a particular department's data is ready. See Monitor console.
- Statistics. Processing statistics can be aggregated by tag.
What tags do not do
Tags have no effect on scheduling. They do not create dependencies, do not influence priority and do not group tasks for the purposes of limits. Use a Group where the intention is to affect execution, and a tag where the intention is only to label.
