Microsoft Fabric : How to take backup of "Direct Lake Semantic Models"

[Binu]

Hi All ,

Have a few questions w.r.t taking backups of "Direct Lake Semantic Models" available in Microsoft Fabric Workspace.

Microsoft Fabric by default is associated to a One Lake Storage(LakeHouse).
1) Can we take backups of the "Direct Lake Semantic Models" in the default Lakehouse or do we have to provision another ADLS Gen 2 just to take backups?
2) If we associate a new ADLS Gen2 with our Microsoft Fabric workspace , will it cause any issue to the default Onelake storage that is already associated with our Fabric workspace. Basically since we are associating two storages with Fabric , will that cause any issue in our Fabric Workspace.

[v-cboorla-msft]

Hi @Binu 

Thanks for using Microsoft Fabric Community.

1) Can we take backups of the "Direct Lake Semantic Models" in the default Lakehouse or do we have to provision another ADLS Gen 2 just to take backups?

• Taking backups of "Direct Lake Semantic Models" By default, Microsoft Fabric is associated with a One Lake Storage (LakeHouse), which is a managed storage service. While Fabric provides versioning and auditing capabilities for your semantic models, it's essential to take regular backups of your models to ensure business continuity and data integrity. Unfortunately, Fabric doesn't natively support taking backups of Direct Lake Semantic Models within the default LakeHouse. To take backups, you'll need to provision another ADLS Gen 2 storage account, which will serve as a backup repository for your semantic models. This approach ensures that your backups are stored separately from your primary storage, providing an additional layer of protection against data loss.

2) If we associate a new ADLS Gen2 with our Microsoft Fabric workspace , will it cause any issue to the default Onelake storage that is already associated with our Fabric workspace. Basically since we are associating two storages with Fabric , will that cause any issue in our Fabric Workspace.

• Associating a new ADLS Gen2 with Microsoft Fabric workspace If you associate a new ADLS Gen2 storage account with your Microsoft Fabric workspace, it won't cause any issues with the default One Lake Storage that's already associated with your Fabric workspace. Fabric is designed to support multiple storage accounts, allowing you to use different storage locations for different purposes, such as backups. When you associate a new ADLS Gen2 storage account with your Fabric workspace, you'll need to configure the storage account settings and permissions accordingly. This will ensure that Fabric can access and write to the new storage account.

Backing Up Direct Lake Semantic Models in Microsoft Fabric:

• Microsoft Fabric's built-in storage (One Lake) doesn't offer native backup for Direct Lake Semantic Models.
• Use an Azure Data Lake Storage Gen2 (ADLS Gen2) account as a dedicated backup repository. This keeps backups isolated from the main storage, adding an extra layer of protection.
• Associating a new ADLS Gen2 account won't affect your existing One Lake storage. Configure permissions properly to allow Fabric to access and write to the new storage account.

I hope this information helps.

Thank you.

[Binu]

Hi @v-cboorla-msft  , Thanks for the information and clarification.

Do you have any link to the above content , if yes can you pls share...

[v-cboorla-msft]

Hi @Binu 

Please refer to the link : How to Backup and restore Power BI Premium semantic models - Power BI | Microsoft Learn

I hope this information helps.

Thank you.

[Binu]

Hi @v-cboorla-msft 

The link that you have shared has no information on Microsoft Fabric , Direct Lake Semantic Models ,"backing up on OneLake" or info on any potential impact/no-impact to fabric workspace if we associate another storage to the workspace apart from the deafult lakehouse.

Could you please share the source/links based on which you had responded to the 2 questions in the original post.

[v-cboorla-msft]

Hi @Binu 

Apologies for the inconvenience.

For now we don't have sufficient information regarding this. The best course of action is to open a support ticket, It's highly recommended that you reach out to our support team.

Please reach out to our support team Microsoft Fabric Support and Status | Microsoft Fabric.

After creating a Support ticket please provide the ticket number as it would help us to track for more information.

Thank you.