Materialization in Snowflake
In Snowflake, Live Query materialization produces results for selected tools that must process the complete input before producing preview results.Live Query reuses the materialized result when you add, configure, or preview downstream tools.
Workflow Refresh
Workflow Refresh refreshes the workflow's source data and marks materialized preview results as outdated. It does not immediately reprocess those results. The next time Live Query needs a preview, it recalculates the affected results using the refreshed source data.
Eligible Tools
The following Snowflake tools and modes support materialization:
Tool or Mode | Why the Preview Limit Provides No Meaningful Benefit |
|---|---|
Sample—First N | The operation must order and number rows before selecting the requested records. |
Sample—Last N | The operation must number the rows and determine the complete row count before selecting the last records. |
Unique | The operation must evaluate every occurrence of each key before identifying the row to retain. |
Record ID | The operation requires an ordered window before record IDs can be assigned. |
Running Total | The operation calculates cumulative values from preceding rows before applying the limit. |
Cross Tab | The operation must aggregate all contributing records before the pivoted rows are correct. |
Create Samples | The operation randomizes the population, assigns positions, and determines the population size before sampling. |
Oversample | The operation calculates class sizes and constructs the balanced result before applying the limit. |
Random Sample | The operation assigns sampling values and calculates row position and total population size before selecting the sample. |
Select Rows—ordered range selection | The operation calculates row positions before applying the range condition. |
Tile | The operation calculates population size, sequence, and tile membership across the complete ordered partition. |
Rolling Window | The operation requires ordered window processing and access to neighboring rows before the limit is applied. |
Selective Materialization
Some tools support materialization only for configurations that must process the complete input before returning preview results.
Tool | Materialize When |
|---|---|
Summarize | The configuration uses operations that must process the complete input before returning results. |
SQL Transform | The SQL contains operations that must process the complete input, group, or ordered partition before returning results. |
Sample | The selected mode requires ordering, row numbering, randomization, or the complete population size, such as Last N or percentage-based sampling. |