Categories
- All Categories
- 76 Oracle Analytics News
- 7 Oracle Analytics Videos
- 14K Oracle Analytics Forums
- 5.2K Oracle Analytics Idea Labs
- Oracle Analytics User Groups
- 40 Oracle Analytics Trainings
- 59 Oracle Analytics Data Visualizations
- 2 Oracle Analytics Data Visualizations Challenge
- 3 Oracle Analytics Career
- 4 Oracle Analytics Industry
- Find Partners
- For Partners
Update data sets built from subject area when subject area changes
We are going through an effort to clean up our subject areas in the Web Semantic Model. Others who started with OBIA likely have the same issue where the out of the box subject areas are less than ideal in 2024 and need to be updated. As we are doing that, we are doing lots of column and folder renames to make it easier for users to understand where to get their data in OAC. For example, the out of the box subject areas give you multiple versions of "Customer" because it could be called Customer, Customer Account Name, Customer Description, or just Cust. We are consolidating these into one dimension called Customer so the user community doesn't need to guess which version to use. By making those changes, it is breaking everything in the data sets and data flows downstream because it's stuck referencing all of the old metadata. We need to be able to update subject areas and have the datasets created from those subject areas also get updated.
I opened an SR with Oracle Support they said that it's "expected behavior" to break everything downstream, which essentially eliminates any continuous improvement opportunities.
Comments
-
Same issue we are facing with simple datasets.
Dataflow architecture is very bad. Every updates we are getting issues.
Dataflows running over year is failing and support team pointing some documents and saying working as expected.We are fed up with regression and new techniques to use everytime.
0 -
If you are renaming and create an alias in the Subject Area — queries will keep working. If there is a case where the alias exists and it is not being used — we should look at that in the SR. If you are removing things which should not exist and were used — you will need to create a list for users to change their workbooks manually with the new column to use.
For the general feature to "replace all references to this column in datasets, data flows, and workbooks", we don't have this feature yet. It is a useful improvement suggestion. LMK if this is the feature you are looking for here.1 -
To your comment about the renaming or aliasing in the subject area, you are correct that the queries will keep working….in the reports. The datasets, on the other hand, do not keep working. That's what sparked my SR and was told that it's expected behavior.
0 -
Got it. The overall improvement is still needed.
Datasets should reference the same column as the alias — unless you use "Edit definition." In this case, it will drop the missing columns. There is an improve to warn about this and help you fix it up. Ping us if you have a dataset that is referencing an alias (without being edited)… we need to investigate this.2 -
Thanks Bret. I have several examples. I started a conversation with Gabrielle. Let me know how I can best help out.
1
