Forum Discussion
Data Bricks in Azure vs AWS
So what are the major differences between Data Bricks in Azure vs the same in AWS. Why should one prefer it in Azure over AWS?
- EricStarker
Community Manager
Thanks for the question! That's our topic for next Tuesday's AMA (AI and Cognitive Services), but we'll have someone take a look at your question either before or on that day.Was this addressed in the session. Could you please share on-demand recording if any?
I could not attend this session... just saw your message
- EricStarker
AliArchitect Hello! An on-demand recording isn't possible given it was entirely text-based - there was no audio or video to record.
You will see that Shunderpooch responded on this thread, who is a Microsoft employee, and so the question was answered.
That AMA was from two years ago, but you can see questions and answers here (sort by oldest thread as this was from October 2020): https://techcommunity.microsoft.com/t5/ai-cognitive-services/bd-p/CognitiveServices
ShunderpoochMicrosoft
ashishkhandelwal2003 There are a lot of reasons I would choose Azure Databricks compared to Databricks on AWS. At a high level, Azure Databricks is a first party service on Azure. What that means is that it's more than a partnership- there are deep integrations between Azure services and Azure Databricks. Examples of what this entails:
- Azure Key Vault Secret Stores in Databricks- Users can work with secrets they have access to, without the secrets ever being exposed to the end user
- Azure Active Directory Support- User credentials on Azure used to authenticate to Azure Databricks, and data in Azure Data Lake Storage accessed through those credentials 'passed through' from Azure Databricks. This allows teams to configure permissions on files to users and groups, and Azure Databricks authenticates the rest
- Top level integration with Azure Data Factory, allowing scheduling of notebooks as jobs in a data estate
- Integrating easily with other Azure Data Services (Cosmos DB, Synapse) through service endpoints on private networks
As a general rule, the integrations to the rest of the Azure platform are deeper on Azure Databricks, compared to how even Databricks on AWS integrates with other AWS services. Overall, this builds a more seamless and streamlined experience for building out your data estate with Databricks.
Thanks for your question! If there's more you want to know, feel free to follow up on this thread!
