Want to create interactive content? It’s easy in Genially!

Get started free

M1-C2 - CDD - Connexion aux sources vraie version

Equipe Data

Created on December 23, 2025

Start designing with a free template

Discover more than 1500 professional designs like these:

Essential Course

Practical Course

Basic Interactive Course

Course 3D Style

Minimal Course

Neodigital CPD Course

Laws and Regulations Course

Transcript

CONNEXION TO SOURCES

Start

Start

Summary

Chapter 2

Will be discussed during this chapter :

  • Introduction to connectors
  • Presentation of the DataPlatform and how to connect to it
  • Connectors good practices in Vicat
  • Differents modes of connexion : Import, DirectQuery, Dual.
  • Connexion template
  • Connexion parameters in Power BI Service
  • Good Practices Recall and takeaways
  • Resources to go further
  • Quiz

OVERVIEW

By the end of this module, you will be able to :

GOAL 1

GOAL 2

Understand all the good practices related to the sources connexion

Setup your own connexion to the DataPlatform

Data Sources introduction

What's a connector ?

A connector is the component that links your Power BI report—and more broadly, your semantic model—to your data source. It establishes a direct data transmission path, enabling Power BI to query and retrieve data seamlessly. Since Power BI is a Microsoft solution, it offers a wide range of built-in connectors to cover diverse business needs. Among these, you’ll find popular options such as SharePoint, Snowflake, SQL Server, Azure services, and many others.

Why are connectors important ?

They simplify integration with multiple data sources. They ensure compatibility with both cloud and on-premises systems. They allow for flexible architecture, supporting Import, DirectQuery, and Dual modes (we'll come back on those topics later in this module).

BEST PRACTICE WARNING

DATA PLATFORM

DATA PORTAL

Data Scientist

Data Analyst

DATA PLATFORM

Datalab

Area where you acts as Citizen Data Developer

EXPO DS SPACE

Data Producer

Data Consumer

Landing Storage(by country)

DWH

Raw Storage(by country)

Internal

Internal

EXPO MODEL

DatabaseRAW

Refined

Incoming

DatabaseCountry

External

Reject

EXPO CDE SPACE

Middleware

Details

Team who take care and automatise Data extraction

DATA portal presentation

As a CDD, you may have needs that can be addressed directly through the data portal. To give you a quick overview, the data portal is the tool that helps members to:

  • Access the data catalog
  • View item lineage
  • Explore various data use cases

Additional tasks can also be performed through the data portal by CDE. For example, if you need a new flow, it can be created directly from this platform. For more information, please contact the CDE responsible for your country.

power bi sources - which one to chose?

A lot of connectors possible...
... that can lead to maintenability issues.
Find out the good practices

power bi sources - which one to chose?

DATA sources - The connectors ranking

To ensure data security, maintainability, and availability, it is essential when creating your reports to prioritize this connector selection logic. Each connector has their own properties, their own advantages and inconvenients. The data connector is the first way to improve the cyber security, the freshness and the quality of the data.

Connexion directly to solutions via connectors known as On-Premises. Often used with a Direct Query mode. It can also involve web connections through APIs. This can be useful in certain use cases, but it lacks the level of data centralization. It's recommended to use the DataPlatform and the Middleware's tools to have more security and be more robust.

Absolutely avoid: using local files (Excel, CSV). It represents security vulnerabilities and prevents scalability and reproducibility. A local file can only be used by a single user. If you want to proceed wih some local files, think of uploading it to a sharepoint folder.

If your data is not (yet) in the data platform, you should then switch to a cloud source, namely SharePoint. SharePoint provides versioning, which offers additional security. Cloud sources are also a good option, as they allow different users to work with the same data source.

Our data platform: Snowflake - Data is mostly reliable and automated. - Data is clean. - The major part of the compute is done in Snowflake to gain in performance. - Data are tabular.

DATA sources - data platform connection

As you may already know, the recommended practice when creating your Power BI reports is to connect directly to the Data Platform: Snowflake. A connector is already available for this purpose. To guide you through this step and ensure the source is properly configured, we have prepared a video (as well as a PDF file) that covers everything you need to know about:

  • Using a connection template
  • Connecting via SSO

The video will be displayed just after this module !

DATA sources - Connexion modes

Which storage mode to choose?

data import - power query

When the source is chosen, you will be automatically redirected to Power Query, a tool that you have already learn the utilization in the Module 2 available on Campus Vicat. We'll not deep dive into Power Query details in this module but we will give you some resources to complete your initial learnings. Indeed, this module will still be oriented on Power Query best practises.

"Documentation is the key to debug "

"A good Data Analyst keeps only the data he needs for his use case"

"Do as much as possible in Power Query...But not everything"

"Choosing the right format at the beginning avoid transformations at the end"

Discover More

Discover More

Discover More

Discover More

This good practice is very important for several reasons :

  • Less weight and so better performance
  • Easier to develop your report
-> Reduce the horizontal cardinality.

This good practice is very important for several reasons :

  • Each format type doesn't have the same weight when being stored.
  • Simplify the operations (you don't have to transform the type if it's already done)
-> Check all your data types in Power Query before importing Data.

This good practice is very important for several reasons :

  • Power Query is a powerful tool but it's still better to do transformations in the Data Platform
  • Some queries such as "merge", "append" have to be avoid.
-> When possible, do the transformations in the Data Platform.

This good practice is very important for several reasons :

  • By adding description to your queries, it's easier to understand
  • It will help the next person to work on the project
-> Comment everything you've done.

Power bi service - after the publication

Let’s now move on to the post-development phase, directly on Power BI Service. Just like Power Query, an initial module already exists, so we will stick purely to best practices. In the next part of the module, various topics will be presented to you, including RLS, apps, and deployment pipelines. Here, we will dive into a technical subject that we will only briefly cover: the use of gateways and the definition of proper connections on the Service once you have completed your development.

Main benefits of gateways and cloud connections

Security

Refresh management

Centralization

A single gateway is being used by many users, facilitating the governance and maintenability.

Provides a secure bridge between data sources and power bi service.

It enables you to set up scheduled refresh to have data frequently updated.

Power bi service - publication & refresh

As said just before, there are several things to know about before publishing & refreshing on Power BI Service.

Publishing a Power BI report / semantic model is something engaging.

2 types of refresh

Discover More

On Demand

Scheduled

Main good practice to know: a maximum of 3 refreshes per day per dataset Too many refreshes can be problematic because we are using shared capacities across the organization. By refreshing too often, you consume a large amount of compute for a long period of time, which can overload the capacity and potentially cause freezing for minutes, hours, or even days if the limits are exceeded for an extended period. This can impact many users. Always ask yourself: do I really need to refresh the semantic model again? If you are already refreshing several times per day and need even more, you may contact the data team to have a specific organization and architecture. Switching to a DirectQuery connector can be a good solution instead of importing data, but it still comes with some drawbacks.

To go further....

  • Clean, Transform, and Load Data in Power BI - Training | Microsoft Learn
  • Get Data in Power BI - Training | Microsoft Learn
  • Design scalable semantic models - Training | Microsoft Learn
  • Access and prepare data for Power BI Desktop - Online workshop - Training | Microsoft Learn

The end

Now that you've learned how to connect properly to data sources and apply all the best practices (choosing a scalable and secure connector, managing cardinality, keeping only useful columns, etc.), it's time to put your knowledge into practice. Go ahead and apply what you've learned by working on your report with reliable sources and accurate data. Good luck, and don't hesitate to reach out to us if needed!
Click here to exit the genially

DATA CATALOG

The purpose of this page is to reference all the items we have in Power BI (reports, dataflows, datasets, apps) as well as in the Snowflake (object names, columns, descriptions).On this page, you will see three sections when viewing an item. Hover over the button beneath the image to explore these features in more detail.

Lineage Window

Sharing Window

Usage Window

DATA & AI APPLICATION

The purpose of this page is to serve as a library of all our Data & AI projects, not only Power BI.However, you will still find a significant number of Power BI reports included.How can CDD use this page?

  • Discover new use cases
  • Get inspiration and guidance from different types of visualizations
  • Get the contact details of the person who created the project

Data source credentials

Sometimes, when you do several modifications you refresh your semantic model and so your report, you will have some error cause by the fact you have to setup again your connections. To do that, you only have to connect yourself with your Microsoft Account and all the dataflows related to your semantic model will be refreshed. It's the same process for the Snowflake connection for example as you only have to configure the cloud connection to choose the gateway you're using.

Gateways

When you publish a report in Power BI Service or add/modify a data source, you may need to redefine the gateways that allow you to connect to those sources. In other words, this step involves setting up the connection you want to use so that refreshes can run properly. For the Data Platform, you will find gateway names as shown in the screenshot above. For other sources such as Dataflows, SharePoint, you will need to include a Personal Cloud Connection, which relies on SSO based on your Microsoft Account.

DATA LINEAGE

The purpose of this page is to give you the ability to retrieve all the sources linked to a specific item.You can search through all the levels available in the data platform. This allows you to clearly see where all your data comes from and where it is stored.

DataPlatform & CDD

When you are a CDD, this is a good practice to connect to the DataPlatform as it ensure reliability, security and scalability. Nevertheless, you won't be working in Snowflake directly. This is a feature used by Citizen Data Engineer. CDE and CDD have to communicate well to optimize their works : from data integration and modelisation in Snowflake to report development in Power BI. It's important to understand Snowflake's structure because CDEs can create schemas to store their views. Please refer to your CDE for more information.

Data source credentials

Sometimes, when you do several modifications you refresh your semantic model and so your report, you will have some error cause by the fact you have to setup again your connections. To do that, you only have to connect yourself with your Microsoft Account and all the dataflows related to your semantic model will be refreshed. It's the same process for the Snowflake connection for example as you only have to configure the cloud connection to choose the gateway you're using.

Gateways

When you publish a report in Power BI Service or add/modify a data source, you may need to redefine the gateways that allow you to connect to those sources. In other words, this step involves setting up the connection you want to use so that refreshes can run properly. For the Data Platform, you will find gateway names as shown in the screenshot above. For other sources such as Dataflows, SharePoint, you will need to include a Personal Cloud Connection, which relies on SSO based on your Microsoft Account.