Lineage Demonstration

Let us walk through an example to demonstrate the power of lineage : Take Apptio One Cost project (historically called as Cost Transparency) and pull up a simple data set in a simple model just to get started.

Consider the Apptio Value Explorer >>Value Master data set. Here, in the data pipeline, you have append step with a couple of tables connected, a formula step including one lookup to another table. You also see Join step which show multiple relationships between tables. These are some of the relationships you can quickly identify here.

To explore lineage for this table, navigate to the project explorer, right-click and select Trace Lineage for this document. A new tab for lineage appears with a fairly complex diagram.

Blue color represents model related objects(AVE) and Tan represents reports.

The initial view is limited to a depth of 1 and will show downstream dependencies as we are starting from a table which can be increased, based on the complexity of the relationships of your selected entity. As the project gets very complex, accordingly you can get more depth to drill down the relationships. You can also switch between downstream dependencies which is default if you pick a table, or upstream dependencies which are default if you pick a report.

You can also include or exclude the sub-components or subtypes associated to the tables, model object or reports. Select the checkbox for the subtype you want to see in your lineage diagram. Numbers with each subtypes denotes the total count of entities present in the diagram. You can also Select all/Deselect all subtypes using toggle subtypes.

Additional information for any entity is available by hovering over them. If you click on the Upside Red arrow, it will show the list of entities to addon above the selected object in the graph.

Similarly, click the Downside Red arrow to see the downstream entities.

Now, remove all the subtypes and select Apply. You can see a table at the top level going into a model. Now increase the depth and apply to see more information with these two entities in focus.

Here, a solid line means there is a direct dependency between the two entities and you can see what that dependency is by hovering over the line. The dotted line means there is an indirect dependency: likely other intermediate entities in between the two on the diagram. You can double click on the dotted line to see what those intermediate dependencies are.

You will also notice that the depth to choose is increased up to level 5, which means we increment just a couple of layers at a time. This is done to retain the same performance with each iteration and return more information. On choosing the maximum depth, you will notice that the table is associated with two different models, which are providing information to multiple reports.

Now that you have some understanding how Lineage works, lets go to another use case where you wanted to look at a column and understand the impact if we change it.

Let us take Benchmark Tower Spend Transform table, and assume we are interested in a particular column % Tower Spend. You can right click on the column and click on Trace Lineage for this column.

In the downstream dependency, it is clear that this metric column is associated with the table and the report known as Benchmarking Summary. On increasing the depth to maximum, we see the following:

Hence, this view gives us information about what all reports and tables the column is associated with. You may also want to see what are the entities his column is dependant on (upstream) and can switch to "The focused entity depends on" in the right panel. Overall, this will help you to identify the impact to get rid of this column.