Replace AzureSQL db with Synapse Datawarehouse with Partitioning/Sharding technique - Possibilities?

Question

Hi, We have SAAS applications that uses Azure SQL as datastore and connects using EntityFramework. We are exploring Synapse datawarehouse to replace AzureSql. We have below questions that need some assistance.

• Can we able to connect to datawarehouse using EntityFramework (with possibility of having schema changes applied)?
• Can we have some sort of paritioning/sharding technique that can be applied in datawarehouse for tenant specific data (using specific identifier value). Once we no longer want specific data we can truncate the data/delete partition associated with identifier.

 

Need some suggestions whether there is such possibility and pl point some documentation. Thank you

Answer 1

@Gowtham Chittoor Welcome to Microsoft Q&A forum and thanks for reaching out here.

Can we able to connect to datawarehouse using EntityFramework (with possibility of having schema changes applied)?

As of today, Azure Synapse doesn't support entity framework. Entity Framework (and ORMs in general) are an anti-pattern for a data warehouse. They're generally designed to work with singleton (or low volume) inserts, updates and deletes.

Here is an existing feature request feedback logged by a different user - Support for Entity Framework Feel free to up-vote and comment on it to help increase visibility.

I would highly recommend to go through this article to get an idea of compatibility checks: Azure Synapse Compatibility Checks

Can we have some sort of paritioning/sharding technique that can be applied in datawarehouse for tenant specific data (using specific identifier value). Once we no longer want specific data we can truncate the data/delete partition associated with identifier.

Yes, Synapse SQL supports table partitions. You can find more info in the below documentation.

1. Synapse datawarehouse - What are table partitions?

Hope this info helps.

---

Please don’t forget to Accept Answer and Yes for "was this answer helpful" wherever the information provided helps you, this can be beneficial to other community members.