This series is intended to be a personal study guide. Information may not be comprehensive or accurate. I am sharing it in case others find it useful. Please feel free to comment if any information is inaccurate.
1.1 Explain how to manage Snowflake accounts
Managing Your Snowflake Account
-
- ACCOUNT_USAGE and READER_ACCOUNT_USAGE schemas provide object metadata and historical usage data about the account and all reader accounts associated with it
- ACCOUNT_USAGE
- In general, these mirror corresponding views/functions in the information schema, with the following differences:
- AU shows dropped objects, IS does not
- AU has a latency of 45 min to 3 hrs, IS has no latency
- AU retains historical usage for 1 year, IS for 7 days to 6 months
- Account Usage views
-
READER_ACCOUNT_USAGE views (estimated latency)
- LOGIN_HISTORY (2 hrs)
- QUERY_HISTORY (45 min)
- RESOURCE_MONITORS (2 hrs)
- STORAGE_USAGE (2 hrs)
- WAREHOUSE_METERING_HISTORY (3 hrs)
-
- System-defined views and table functions that provide metadata about account objects
- Based on SQL-92 ANSI, but with additional views and functions specific to Snowflake
- ANSI uses 'catalog' to refer to databases, Snowflake does the same to stay compatible
- Snowflake automatically creates INFORMATION_SCHEMA in every database
- The schema contains the following:
- Views for all objects in the database
- Views for account-level objects
- Table functions for historical and usage data across account
- INFORMATION_SCHEMA is read-only
- Queries on INFORMATION_SCHEMA do not guarantee consistency with concurrent DDL (objects created during the execution of the query may or may not appear)
- Output depends on privileges of user's current role
- Views are optimized to return small subsets of data, use filtering whenever possible
-
- Costs are based on usage of stored data, virtual warehouses, and cloud services
-
Data storage
- Usage for data storage is calculated on the daily average amount of data (bytes) in:
- Staging
- Database tables
- Fail-safes
- Clones of database tables
- Monthly cost is based on a flat rate per terabyte, type of account, and region
- Usage for data storage is calculated on the daily average amount of data (bytes) in:
-
Virtual warehouses (compute)
- Usage for virtual warehouses (compute) is calculated based on Snowflake credits consumed for:
- Executing queries
- Loading/unloading data
- Performing other DML operations
- Credits are charged based on number of VWs, their size, and how long they run
- VWs come in ten sizes, each twice the power and cost of the previous
- VWs are billed only when running, not when suspended
- Credits are billed per second with a 60s minimum
- Usage for virtual warehouses (compute) is calculated based on Snowflake credits consumed for:
-
Cloud services
- Usage for cloud services is charged only if the daily consumption of cloud services is greater than 10% of the daily usage of compute
Top comments (0)