When your business delivers analytics as a service, a critical component of your infrastructure is the data warehouse. It’s what drives the analytics interface your customers depend on. Whether your customers are launching queries for generating dashboards, filtering audience views, or generating BI reports, a constant challenge is scaling your data infrastructure without hindering your services or stalling your customers.
Data growth continues to explode
Given the insatiable appetite of businesses to analyze more and more data, the opportunities for your analytics services will increase too. Is your data warehouse ready to handle rapid data growth, year-over-year, for potentially thousands of customers?
If your analytics service is aligned with a fast-growing industry segment, it’s reasonable to speculate that some organizations will grow as much as 200 to 300 percent a year. This could amount to 10X data growth and data processing in just a few years. Two hundred terabytes of data can quickly become two petabytes.
Build or buy the infrastructure?
If you concluded that immense scale is something worth planning for, the next obvious question is, “how do I get there?” If your analytics service is built with on-premises data warehouses, continually adding racks of hardware is not only expensive, it will have diminishing returns as you run out of physical data center space. This is why you see state-of-the-art, 2.5-in. form factor disk drives approaching 15TBs of capacity, per drive. Despite increasing disk drive capacities, swapping out hardware will remain expensive and disruptive because it involves downtime. And, the addition of racks doesn’t resolve other important concerns, such as data loading, backup, data protection, and security.
If you push your data warehouse infrastructure to the cloud, you’ll have the same issues. The difference is that you’re now managing cloud instances. Although this is faster than procuring hardware, it’s complex and you’re still left expending resources to build and manage the infrastructure. Behind the scenes, your data team will constantly work around issues as they attempt to prevent bogging down your analytics services as they expand cloud infrastructure.
Instead of these options, consider a ready-made data warehouse that’s built for the cloud and delivered as a complete service that you can connect to and use to develop your analytics application services.
The opposite of scaling quickly is operating slowly
Using a cloud data warehouse that’s specifically designed to infinitely scale, without disruption, provides you the best performance footing. Customers come to you because they want to run analytics faster. If you can’t speed up a query, there’s a domino effect of negative consequences.
Accelerating service provider analytics
Snowflake provides analytic service providers a complete, ready-made data warehouse infrastructure that’s engineered to scale elastically and automatically, with no infrastructure to manage.
You can develop your customer-facing analytics to connect to Snowflake with ODBC- or JDBC-driven integration. You can also create as many independent virtual warehouses (compute engines) as you need to support your customers (see Figure 1). You have complete control over sizing compute resources for each virtual warehouse.
After setting your configuration, you have the option to allow the Snowflake service to scale itself and bring a new group of resources online that match the configuration. The benefit to you, and ultimately to your customers, is that Snowflake’s virtual warehouses instantly and automatically scale. No database administrator is required, thus eliminating workflow bottlenecks that expose you to slower business execution.
Snowflake Cloud-built Data Warehouse
Data loading options
You have the flexibility to choose your method for loading data into Snowflake using Snowpipe, our data ingestion tool that automatically loads data from Amazon S3 or Azure ADLS. Or, use one of the many Snowflake data-integration tools available from our partners.
Load structured and semi-structured data such as JSON data on behalf of your customers, all within the same data warehouse. Your customers can execute complex queries, including joins, without performing any pre-transformations.
SPEED QUERY EXECUTION, WHILE KEEPING COSTS DOWN
Snowflake’s results cache accelerates performance for identical queries against the exact same data set, and query results are pulled directly from cache without generating any compute charges.
With built-in security features, Snowflake ensures data protection, processing failover, end-to-end data encryption (and more), with a FedRAMP In-Process platform that is HIPPA, SOC 2 Type II, and PCI DSS compliant.
After loading your data, you’re ready to orchestrate data access with Snowflake’s built-in SQL tool, with Python, R, or other supported data access and development tools. For supporting your internal test-development efforts or customer support, simply clone any size production database on-the-fly using Snowflake’s zero-copy cloning tool.
It all adds up to helping you maintain the fastest execution possible to meet the growing demands of your customers. With Snowflake, you can allocate more of your valuable data engineering resources to creating new products that advance your analytics service and keep you ahead of the competition.
The post How Snowflake Accelerates Analytics for SaaS Developers appeared first on Snowflake.