General Introduction
Basic ideas about Snowflake
- Separation of storage and compute
- Built for Cloud environment
- built from scratch
- optimized for cloud
- storage & compute is decoupled
- Software as a Service
- No software, infrastructure or upgrade to manage
- Pay only for used compute & storage
- Storage & compute charged independently, only for use
- Scalable
- Virtual warehouse enable compute scaling
- Authentication
- Configuration
- Resource Management
- Data Protection
- Availability
- Optimization
Unique features in Snowflake
- Time travel
- Cloning
More details can be found at {% post_link snowflake-online-talk ‘Snowflake Online Talk’ %}
Basic components in Snowflake
- Databases
- data storage layer
- Warehouse
- data computation layer
- Worksheet
- query window
Price Model
Storage Cost
- AWS S3 or Azure Blob Storage
- Based on actual usage
- Columnar Compression (e.g. {% post_link all-about-parquet ‘Apache Parquet’ %}) or other techniques to save cost
Compute Cost
- Loading data
- Processing data
- Only for actual usage
- Virtual warehouse can help minimize costs
Data Transfer
- No cost for transferring into Snowflake
- Costs on transfer out to other regions and to other cloud platforms
Metadata
- no cost
- Database definitions
- table definitions
- users permission configuration
