What Are Hybrid Tables in Snowflake?

During the 2022 Snowflake Summit in June, Snowflake announced a new table type that supports its Unison workload, a table type referred to as a Hybrid table.

It’s been in the data news and publications for a while now…So what’s all the hype about?

To Know more about these trending topics and to leverage these game-changers within your environment, stick around until the end!

In this post, we will explain the concept of the Unison workloads, hybrid tables, their advantages, and a typical UniStore use case.

What is Unistore?

Today’s companies constantly generate transactional data from all parts of their business. 

Typically, this data is siloed across a multitude of legacy databases, and analyzing this data requires replication into some form of data repository. 

This historical division has existed for decades, and companies like SAP have taken a shot at bridging this gap with solutions like their HANA database, which stores data in column-based tables in main memory and brings online analytical processing (OLAP) and online transactional processing (OLTP) together.

Similarly, Snowflake, the data cloud platform, has introduced the Unison  workload, a modern approach to using transactional and analytical data together. 

What are Hybrid Tables?

Hybrid tables, also known as Snowflake Unistore hybrid tables, underpin Unistore. These tables are optimized to support analytical and transactional workloads concurrently.

The Syntax for creating hybrid tables is not so different from the Syntax used for normal tables, with a few exceptions, namely:

• A Hybrid table must have a primary key constraint
• Must have unique constraints for building underlying index

See the snippet below:

As you can see from the image above, it resembles a typical DDL query, all-inclusive of those exceptions.

Since the hybrid tables are a type of snowflake table, it’s a no-brainer that they also inherit the capabilities of a typical snowflake table; properties such as time travel, data masking, row-level locking, etc. are applicable to hybrid tables.

The Advantages of The Unistore Approach To Data

Fewer Systems To Manage:   With all your data in Snowflake, there’s no need to manage your application, server, and UI layer in different environments. 

The Unistore approach simplifies and streamlines enterprise app development directly on Snowflake.

Act-On Data Immediately:  No more Delayed Access. With the Unistore approach, we can get powerful insights into our businesses by running analytics on transactions as they happen in real-time.

Eliminate Data Movement:  No more redundant and replicated data sets.

Standardize Security and Governance Controls:  With features like tagging and dynamic data masking, we can say goodbye to data Governance and security challenges.

A Typical Unistore Use-Case

Ps: At the time of writing, This feature is not yet publicly available (as shown below).

Attempt to use hybrid feature in Snowflake environment

To get notified when the public preview is available, please sign up  here .

However, If you’d like to see these features in a demo,   please check out the video below.

In this demo, Carl Perry, the director of Product management at Snowflake, reveals how Unistore and Hybrid Tables can power an internal business application and get immediate insights into sales and operations.

Wrap Up

If you already use Snowflake, the Unistore and hybrid table feature is definitely worth being excited about. 

You’re maybe even starting to imagine how you can manage your apps and perform analysis on your transactional data concurrently…it holds so much potential, right? 

Definitely 💯.

The bad news is, getting insights on our transactional data would still involve long lines of complex query code.

But as you can expect, we come bearing good news! 😎!

What if I told you your data modeling process doesn’t need to be full of nested subqueries and complex lines of SQL code? 

With no-code modeling tools like Datameer , you can say goodbye to all of that.

With our easy drag and drop modeling features, you can shorten the time between query-building and generating real-time insights.

Interested?

Learn more by visiting the Datameer website , or get started immediately with a free trial of our multi-persona platform.