getting started in fabric and DF alternatives for etl

[db042190]

Hi I'd like to set up a fabric alternative to our ssis based data warehouse etl.   What would i need to get started?  Is it called a fabric license or maybe capacity?   nobody else needs it except me for development and then a way to run it in production.  Will tyhe look and feel be kind of like power apps demos i've seen or something else?

I've heard Data Factory (not calling ssis from DF) might also be an alternative especially if we are thinking about big data down the road.  I heard MS may have opened things up a bit in DF where you dont have to be big data to use it instead of fabric for etl alternatives.   Is that a license or capacity?   i would only need it for developemnt and then a way to run it in production.

I think on both understanding context would help, ie are they associated with me, a workspace or something else?

[V-yubandi-msft]

Hi @db042190 ,

Thank you for engaging with the Microsoft Fabric Community. Just to add a bit more context from my side in case it helps others evaluating Fabric vs Azure Data Factory for SSIS replacement. Fabric provides a unified platform combining lakehouse, pipelines, and Power BI. To use it, you'll need a Power BI Pro license and a Fabric capacity F2 is a suitable option for development.

Alternatively, Azure Data Factory is a strong choice if you want Azure native orchestration. It operates on a pay per use model and doesn't require capacity, making it practical even without large data needs.

For development and production environments, you can set up separate workspaces with different capacities or use deployment pipelines to move artifacts between them.

Thanks for your response @tayloramy @GeraldGEmerick .

[db042190]

thx v-yubandi.  on the fabric answer what if we dont have lakehouse?  just a sql server warehouse being loaded from various sql and non sql connectors?

for adf i must need something.  at least a non pro license?  can it load a sql warehouse from various sql and non sql connectors?

most importantly whats the difference between a unified platform and Azure Native Orchestration?  maybe a sentence or 2 would suffice.

[V-yubandi-msft]

Hello @db042190 ,
You don’t need a lakehouse to use Fabric. Pipelines (Data Factory inside Fabric) can move data straight into your SQL Server warehouse, similar to SSIS. The lakehouse is just another storage option if you want it.

With Azure Data Factory, there’s no Pro or Fabric license required, it’s pay-per-use and supports loading SQL Server from many SQL and non-SQL connectors.

The big difference is: Fabric = everything in one place (data movement, storage, reporting with Power BI), while ADF = Azure-native service focused on pipelines/orchestration only.

Regards,
Yugandhar.

[tayloramy]

Hi @db042190,

The first thing you need is a Fabric Capacity, you could start at an F2 and then scale up as @GeraldGEmerick mentioned. 

You will also need a Power BI Pro license to be able to create fabric objects. If your org has E5 licensing this is included, but may need to be enabled on your account depending on your set up. 

The basic overview of how Fabric is structured is as follows: 
Capacity - this is your compute. 

Workspace - this is like an SSIS project - a place to put things. Workspaces are assigned to a Capacity. 

Artifacts - these are the individual items - data pipelines, copy jobs, notebooks, dataflows, lakehouses, ect. These exist inside a workspace. 

If you found this helpful, consider giving some Kudos. If I answered your question or solved your problem, mark this post as the solution.

[db042190]

thx amy.   i'm starting to catch on.   all employees alreayd have pro licenses.  still not sure if a lakehouse is something created inside one of these flows or an already existing (maybe large) data store.   will look up the definition here in a moment.

[tayloramy]

Hi @db042190, 

A Lakehouse is a fabric item, created inside a workspace. 

If you found this helpful, consider giving some Kudos. If I answered your question or solved your problem, mark this post as the solution. 

[GeraldGEmerick]

@db042190 You would want a Fabric capacity. You could start with an F2 and then scale it up later. Each Fabric capacity comes with all the tools you need including data factory and dataflows that you can land in a lakehouse or data warehouse.

[db042190]

thx gerald.   didnt know df comes with.   when working in fabric and trying to move data, what criterion makes one choose insertion of a df object as opposed to using just the other data movement "things" you get with fabric ?   is gen2 part of df or more a part of fabric exclusive of df?  and chicken and egg, had the ws better exist before our entra guys give that capacity?

[GeraldGEmerick]

@tayloramy There are a lot of different ways to move data within Fabric including Dataflow Gen1, Dataflow Gen2, Data Factory, Python notebooks, etc. From what I have seen, Dataflow Gen2 is potentially the fastest while Data Factory tends to be less expensive but can take twice as long as Dataflow Gen2. You'll have to do your own testing though as your data movement may be different than what I have seen others test with.