Critical Success Factors for the Cloud Data Lake
Webinar Speaker: Philip Russom, Senior Research Director for Data Management
A cloud data lake integrated with a cloud data warehouse is an emerging architecture that will soon be prominent. Success of this architecture depends on data lake design and cloud data platform selection.
According to a 2019 TDWI survey, roughly half of data warehouse programs surveyed have a data lake in production, whether deployed on premises, on the cloud, or both. The survey data makes significant points:
- The data lake is firmly established as a method for organizing analytics data
- Data lakes and data warehouses coexist and complement each other
- Cloud is becoming the preferred platform for both lakes and warehouses
This webinar will drill into the above points to help data management professionals understand the critical success factors for a cloud data lake, especially when it is integrated with a cloud data warehouse.
Webinar attendees will learn:
- Why cloud for data lakes and data warehouses? Why now?
- Prominent data lake use cases—including data exploration at scale, self-service data practices, and advanced analytics—and how a cloud data platform must enable these
- Failure factors, such as poor platform choices, inadequate metadata, latency issues, underestimating relational requirements, and letting a lake become a swamp
- The important roles that SQL and the relational paradigm play in a cloud data lake
- Architectures and best practices that unify a cloud data lake with a cloud data warehouse
- Best practices for migrating an on-premises data lake or data warehouse to cloud
- High-priority characteristics of cloud-based data platforms for lakes and warehouses
- How adopting a general cloud data platform can simplify the deployment and raise the business value of a cloud data lake integrated with a cloud data warehouse
- How Snowflake is a cloud-based data lake and data warehouse all in one platform that can index all structured and semistructured data, transform it, and make it available for fast analytics
Philip Russom, Ph.D.