The Clearly Podcast
Azure Synapse Analytics
Summary
This week we discuss Azure Synapse Analytics
This is, in essence, a cloud-based distributed database system. It is aimed at larger scale analytics applications rather than transactional processing.
The approach taken by Synapse to processing queries is to break them down and process parts of a query across different computing nodes. The result sets are returned to a control node where the final result is processed. If, for example, you request an average of a field, the control node will request the various compute nodes to calculate the sums and record counts, that can be returned and recombined into the average originally requested.
It is important to note that the higher performance in processing analytical queries tends to come at the expense of concurrency - but in an analytical application, this tends to be less of an issue.
Finally, we should also note that given the price point of Synapse, any company embarking on a data analytics journey with any expectation that their data may get large, should seriously consider Synapse as part of the tool set they are considering.
More about Synapse can be found here.
If you already use Power BI, or are considering it, we strongly recommend you join your local Power BI user group here.
Transcript
Andy: Gentlemen, how are we doing today?
Tom: Yeah, I'm good, thanks. How are you?
Shailan: All good.
Andy: Excellent. Today, we're excited to talk about Azure Synapse Analytics—a limitless analytics service that integrates data warehousing and big data analytics, offering both serverless and dedicated resourcing. Ready to dive in?
Shailan: You've been into the marketing pitch again?
Tom: So today, we’re talking about Microsoft Access database?
Andy: No, we're actually talking about Azure Synapse Analytics. We'll explain what it is, who it's for, and when to use it.
Tom: Synapse Analytics is essentially a distributed SQL Server database system. It spreads compute resources across multiple CPUs, enabling you to handle much larger datasets than traditional databases. This concept dates back to Microsoft's parallel data warehouse appliance. Now, with the cloud, compute resources can scale up and down easily, reducing costs. This makes it accessible to medium-sized businesses with large datasets—tens to hundreds of millions of rows or more.
Andy: So, this technology might trickle down to smaller businesses over time. Do you see that happening?
Tom: Yes, the branding shift from Azure SQL Data Warehouse to Synapse Analytics reflects its broader usage. Even businesses with around 100GB of data might find it cost-effective.
Andy: So, mid-sized retailers with significant transaction volumes could use this?
Tom: Absolutely. It’s no longer just for large companies; medium-sized businesses can also benefit from it.
Andy: Where is Synapse effective, and where is it less useful?
Tom: Synapse is great for analytics, not for transactional reporting. It excels at handling large datasets and complex calculations. It's not meant for serving up small data bits to many users simultaneously.
Shailan: Exactly. It's built on SQL, so it's familiar to many. It's also consumption-based, making it cost-effective. Plus, it supports various languages like T-SQL, .NET, Python, and Spark, catering to different data analysis needs. Its connectors make it versatile and feasible for many organizations.
Andy: Microsoft is investing heavily in Synapse, which is reassuring for users considering this technology. It indicates long-term support and development.
Tom: Yes, it's built on the Azure platform and leverages SQL's scalability. While there are some differences, those familiar with SQL can quickly adapt.
Shailan: For our next podcast, let's discuss transitioning skills from traditional data warehousing to cloud solutions like Azure Synapse.
Andy: Great idea! Staying updated with new technologies is crucial for any specialist.
Andy: Alright, to wrap up, we don't have specific top tips today. Instead, we explained what Synapse is and its use cases. If anyone needs help, visit clearlysolutions.net or clearly.co.uk. Microsoft has excellent resources beyond the marketing fluff. We'll also discuss skill transitions in a future episode.
Andy: Thanks, everyone!
Shailan: Thanks a lot.
Tom: Bye, guys.