To provide complete analytics solution Microsoft combined different services into one service, and that's how Azure Synapse Analytics evolved. Azure Synapse Analytics is basically a workspace, from within all different services can be accessed seamlessly without network or security issues. So basically in Synapse Analytics you can have the Power BI, you can have Dynamics, machine learning, data store, ETL, tools like Data Factory, Big Data, everything combined into one service including data warehouse.
So how does the Azure Synapse workspace look like in the Azure portal ? See below
The above figure shows the MPP architecture which Synapse Analytics service uses. MPP stands for massive parallel processing. As there is a need to process terabytes of data, millions and billions of rows of data, a system is required which can parallelly process these records. MPP has the Control Node where application submit the queries. The MPP engine actually runs inside this Control Node which distribute the data to all the other Compute Nodes. These Compute Nodes can be considered as the worker nodes. They also exchange data between each other using DMS, data movement service. And finally, after processing the data within these Compute Nodes, all these Compute Nodes return back data to the Control Nodes. And finally, Control Nodes return data back to the application.
So control is always in constant communication to the different compute nodes that are part of the system. The control node will always be the same regardless of how many data warehousing units you add. However, you will get more compute nodes as your DWUs increases.
Azure Synapse Analytics is a fast, flexible, and trusted cloud data warehouse that let's us scale not only our compute, but also our data store. We can scale them elastically and independently of each other so we can scale up our compute services and leave our data store size alone, or we can scale in both up, we can scale in both down. If we notice that we're not using as much storage, we can scale it down and leave to compute where it is, so again, independently. With a massively parallel processing architecture, Synapse SQL leverage is a scale-out architecture, to distribute computational processing data across multiple nodes. Compute is separate from storage, which enable to scale compute, again, independently of the data in your system.
Dedicated SQL pool, the unit of scale is an abstraction of compute power and also known as a Data Warehouse unit.
For serverless SQL pool, scaling again is done automatically. Serverless implies that Microsoft manages and allocates the compute power of this SQL pool. Also there is no infrastructure to setup or clusters to maintain.
The Serverless SQL Pool is created simultaneously when your create your Synapse Workspace resource in Azure. For each Synapse workspace there is a default built-in endpoint for a Serverless SQL database. It’s a node-based design based on the DDP system, or Distributed Data Processing system. In this system query’s are split into smaller query’s, and executed on the compute nodes. The control node on top utilized the distributed query engine to optimize the query’s for parallel processing. Each small query is called a “task” and represents a distributed execution unit.
The Dedicated SQL Pool can be put on pause to save costs, and the Serverless Pool can take over to create views or external tables for querying, and afterwards it can be imported into Power BI.
Azure Synapse Analytics supports both SQL Server Authentication and Azure Active Directory.
For high security environments, setup multifactor authentication.
From a data perspective, Azure Synapse Analytics Support security at the level of both columns and rows.
Reference and Credits :
Microsoft.com
Comments
Post a Comment