Base Bi - MYSQL

Asked

Viewed 35 times

-2

It’s a layman’s question for being a situation I hadn’t encountered before. I have an instance of RDS Mysql with at least 22 databases. The current company needs to build a bi-model to consume the data from these databases. Having "multiple Ids", being one for each customer, is impracticable due to cost. Which would be the best option:

  • Build a large database with the data of all customers and on this basis map which data is from which client?
  • Export Excel files with the data of each client and the Power bi consume the data?

Would you have any other suggestions? How and what would be the best way to do this?

As I reported is a new situation and had never gone through it. Soon it is completely a layman’s doubt.

  • Nathann, good morning! If the databases already exist, what would be the gain in creating 'a bigger one'? I’m asking just to understand your case. Hug!

  • The production server is in a closed environment and the BI will be allocated to another server. The production bases will be restored daily to a secondary server and the BI will consume the data from that instance for security. Another point is that the responsible for BI informed that the Power BI software does not make multiple database connections. Hence the need to either create a larger base or export files via Excel.

1 answer

0


I’m not sure the question is very clear. Are the 22 databases exclusively customer information? They have the same scope?

The possible answers depend on the objective of the company and the answer to the above question. If they are 22 bases of the same scope, that is, all dealing with similar customer information, the unification of these bases can be done in two ways, depending on whether or not the reuse of the bases as separate or unified instances.

This unification can be done via Excel, as commented, and imported into Power BI as a single source; or, this unification can be done directly in Power BI, through the tool "Transform Data" -> "Add queries".

As stated earlier, the approach taken depends on the reuse or not of these bases as a whole or as separate sources, avoiding re-work (since the product is the same for both approaches).

Now, if the answer to the question raised regarding the content of these bases explains that they are distinct scopes among themselves, I suggest taking a look at the concept of Data Warehouses.

  • Databases have the same database structure (tables, views, procedures, functions), but the data itself differ between clients. Mainly for dealing with customers from all over Brazil. I will take a look at this concept of Data Warehouse. Helped a lot, thanks!

  • Just making an addendum: if the company needs to compare information from different bases, I believe the Power BI nay make these separate connections via Direct Query. Therefore, one of the two approaches would be necessary: to unite everything in a single database or perform the work of importing first into Excel and then into Power BI. If this relationship between different bases is really necessary, I suggest an analysis on the possibility of restructuring the base, aiming to optimize the BI process from a larger work in the base itself.

Browser other questions tagged

You are not signed in. Login or sign up in order to post.