Is it better a single dataset or one for each report?

[fabiomanniti]

What I'm asking here is not really a help but more a tip or suggestion:

For my company I make reports on PBI and I take data from a PostgreSQL datamart.

For what is your experience, do you think it is better to create one big dataset with all the tables in it and then create reports from that dataset or create a smaller dataset for each report I must create?

Obviously there are pros and cons for each choice but I still cannot decide what is better and I was hoping someone chould help me to see it more clear.

Having one "big" dataset:
Pros: 

• If you need to edit something, you only have to do it once
• One big dataset is lighter than the sum of many small one because there are some tables that you will use many times
• You only must update one
• All reports are faster (I'm not sure about this. I'm just guessing)

Cons:

• If you "lose" it, you lose all reports underneath it
• The file itself is heavier so working on it might be frustrating if you don't have a computer with very high CPU and RAM
• Managing connections can be though because you can find a conflict with some table that you don't even need in that single report

To your experience, what do you find better?
And do you have any more pros and cons? 

So that I can evaluate better my choice.

Thank you so much

[v-henryk-mstf]

Hi @fabiomanniti ,

Whether the advice given by @DataInsights  has solved your confusion, if the problem has been solved you can mark the reply for the standard answer to help the other members find it more quickly.

If not, please point it out. I will answer you as soon as possible.

Looking forward to your feedback.

Best Regards,
Henry

[DataInsights · Accepted solution]

@fabiomanniti,

These are good questions. Here are some things you might consider.

Create a golden dataset for each topic area (Finance, HR, etc.). Publish these reports to Power BI Service. These reports don’t need visuals; the dataset is the important part. Using these golden datasets as the source, create reports with visuals (thin reports). Use Power BI Desktop to create thin reports, in order to avoid losing the report if the golden dataset gets deleted (see link below regarding deletion of datasets).

Use dataflows in Power BI Service. A dataflow is Power Query online, and can be used across datasets (single source of truth).

https://exceleratorbi.com.au/new-power-bi-reports-golden-dataset/ 

https://nisorness.com/data/2020/11/19/how-to-prevent-the-deletion-of-datasets-in-power-bi