The Sirius Data & Analytics Consulting team recently attended Snowflake Summit 2022 in Las Vegas; the first time the annual conference has been held in person since 2019. This year’s event was a departure from the much smaller inaugural Snowflake Summit 2019 held in San Francisco. Whether it was due to being in a room full of data enthusiasts or the magic of Las Vegas, the energy matched the larger attendance and venue.
Top Snowflake Summit 2022 Announcements
I would like to highlight a few of the many announcements from Summit that our team is most excited about. While some of the new features expand Snowflake’s capabilities, there are many that support migrations, unified workloads, and hybrid architectures. So, let us look at our favorite updates.
1. Apache Iceberg Support
Apache Iceberg is a high-performance file format that is often used in large-scale distributed systems such as Apache Spark. The advantage of Iceberg over other file formats is its support for expressive SQL, schema evolution, and overall performance. Additionally, Iceberg is supported by Spark, Apache Flink, Presto, Apache Hive, Apache Impala, and more. Snowflake adding support expands interoperability with data lakes and other toolsets. At Sirius, we often see external tables for data workload migration to Snowflake or accessing data produced by other pipelines or platforms. It’s refreshing to see Snowflake’s support for Iceberg and open formats.
2. On-Premises External Tables
External tables are not new to Snowflake, and are a common object in most modern databases. However, adding support for Snowflake to access on-premises storage solutions is transformative for the platform. Any storage device that is Amazon S3 compatible will now be accessible by Snowflake using external tables. This includes Dell ECS, Pure Storage, and MinIO.
3. Tag-Based Data Masking
Data governance is critical to any successful data program. Snowflake’s new tag-based masking policies add a much-needed functionality change that enables easier use, future capabilities, and less overall code to maintain. Policy administrators can create a policy for each data type and assign all of those policies to a single tag. Once the tag is applied at the schema level or higher, all base table columns are protected, enabling sensitive data to be protected while retaining analytical value. This is just one of many announcements that our data governance team is looking forward to working with clients to implement.
4. Native Application Framework
The Native Application Framework is a combination of core Snowflake features and new functionality that helps enterprises build, deploy and monetize applications all within Snowflake. Having an entire application run in Snowflake allows clients to reduce costs, improve governance, and break down data silos. Imagine using the Snowflake Data Marketplace and leveraging an identity resolution app directly in Snowflake, and then enhancing it with a custom recommendation engine application built in-house. The possibilities are endless.
5. Snowpark for Python
While Snowpark for Python has been in the works for some time, Snowflake has officially announced that Python support is in public preview, making the feature available for all Snowflake customers. In addition, we get Snowflake Worksheets for Python as well as integration with Anaconda as the go-to package manager. The combination of features uniquely positions Snowflake to tackle a slew of new workloads, including machine learning. Data engineers, data scientists, and application developers now have a flexible environment to build modern and scalable data platforms.
6. Unistore and Hybrid Tables
This feature was the biggest surprise, and quite honestly we feel it’s a real game-changer. The intention of Unistore—and specifically Hybrid Tables—is to allow Snowflake to tackle online transaction processing (OLTP) workloads. Hybrid Tables are comparable to traditional tables in Snowflake but focus on point lookup query performance using primary key constraints and underlying indexes. In combination with Snowflake’s Native Application Framework and Snowpark for Python, we start to see a focus on reducing overall complexity and data silos. Snowflake has proven its scale and performance with analytic workloads, and we are excited to see the impact made in the transactional world of data.
Remember, this is not an exhaustive list of Snowflake Summit 2022 announcements. For the full list, please check out Snowflake’s Inside the Data Cloud. What new features are you most excited about? If you are new to Snowflake and need a partner to help navigate the data cloud, Sirius has a wide range of data architects, engineers, analysts, and data scientists focused on how to best use and monetize data throughout your data estate.
As a Snowflake Premier partner and a top integrator of modern data platforms, Sirius has a unique perspective on data platform life cycles and maturity. Learn more about our Snowflake Data Cloud services or contact us for an assessment of your Snowflake environment or data and analytics architecture.