2020 was a great year for Snowflake—because we equipped users from data analysts to IT teams with an ever-growing array of tools. Here’s a look at the news and insights that the Snowflake blog community (voting by clicking, of course) read the most in recent months starting from number nine and working up to number one.
9. Reducing ETL friction
AWS Glue gives developers an easy way to build and manage tools for data extraction, transformation, and loading (ETL). Sounds like a natural fit with Snowflake—and indeed, this post gives the why and the how, including technical instructions and a sample ETL script.
8. Real-time analytics with Apache Arrow
Early in 2020, we released new versions of the Snowflake JDBC and Python clients to return results in Apache Arrow’s standard columnar format. The result is less overhead and faster results when Arrow fetches data from Snowflake. This post provides performance benchmarks as well as instructions for getting started with those Snowflake clients.
You’ll notice a theme across many posts last year: The Snowflake team is constantly improving Snowflake’s ability to work with other products and partners across the industry.
7. Working with Azure Data Factory
In this roll-up-your-sleeves post, Snowflake Sales Engineer Jeremiah Hansen details “a better way to connect Snowflake to Azure Data Factory.” His write-up includes step-by-step instructions and a link to the Snowflake-ADF connector repository on GitHub.
6. Single sign-on security
IT teams know reliable single sign-on (SSO) capability makes users more productive.
Here, Vikas Jain, Snowflake’s Product Management Leader for Security Features, walks through Snowflake’s SSO integration with Microsoft’s Power BI. “It solves an important use case for joint customers to integrate their identity provider (IdP) for authentication, such as Azure AD (AAD), Okta, and others, while providing a seamless SSO experience,” he wrote.
5. A steady cadence of new features
As organizations collect more data, their appetite for insights rises right alongside. Snowflake’s product teams kept busy in 2020, adding new features and functions to open up new possibilities.
In this post, Christian Kleinerman, Snowflake’s VP of Product, rounds up some of those innovations and advances along the product roadmap. He groups advances into three categories:
- Core platform features, such as Snowsight, which improves the data analyst experience, fostering more productivity and better collaboration
- Extensible data pipelines, such as partition-aware exports to work more easily with external tools and support for Java-based pipelines and transformations
- And Data Cloud content, such as the fast-growing Snowflake Data Marketplace and Snowflake’s partnership with Salesforce.
4. Customers first, always
We’re proud to be customer focused, first and foremost, but it’s also nice to take notice when others recognize the increasing value and power of Snowflake. In 2019, Snowflake emerged as a Leader in Gartner’s Magic Quadrant for Data Management Solutions for Analytics.
This post also offers some insight into Snowflake’s excellent Net Promoter Score, reflecting Snowflake’s customer-first approach.
3. Snowflake on Azure
It’s a multi-cloud world. This post (from 2018, and still popular) discusses how Snowflake can take advantage of Microsoft Azure features such as:
- Limitless storage accounts
- Accelerated networking
- And storage soft delete
2. Supporting different data models
Many businesses have different data tucked away in a variety of data models. Snowflake works equally well across many different data modeling schemas.
This article from Kent Graziano, Snowflake’s Chief Technical Evangelist, delves into the details of a particular model, Data Vault, that a number of Snowflake customers have asked about.
…And the most popular Snowflake blog post in recent months was:
1. Pricing with no surprises
There’s no question that cloud computing has changed the business world. For many cloud users, though, predictable costs have turned out to be a challenge (to express it gently).
The most-read blog post last year explains how simple predictable costs can truly be. Today, a Snowflake bill consists of only three charges: how much storage you used, how much compute you used, and data transfer costs. There are no hidden costs, no idle compute charges, and no extra fees.