Datashare Schema Reference

This reference describes how Unqork module data is structured in Snowflake. Use it to understand the table structure, how data is stored, and sync behavior.

Tables

Datashare provides up to five Snowflake tables, one for each record type. Each table contains data across all modules in the environment.

Data Format

All data in each table is delivered in raw JSON format. Each record is stored in a single VARIANT column named RAW_DATA. Data is not flattened into individual columns. All component values, metadata, and system fields are accessible as properties in RAW_DATA.

The column structure does not change after Datashare is configured. Unqork does not add, remove, or modify columns when module components change.

For example, tabular data from Grid and Repeater components is stored as a nested array in the data field of RAW_DATA. The array is grouped under the component's Property ID. A Grid component with the Property ID myTableData and columns colOne and colTwo stores its rows as follows:

{
  "data": {
    "myTableData": [
      {
        "colOne": "First",
        "colTwo": "Second"
      }
    ]
  }
}

Sync Frequency

Data syncs to Snowflake on a default one-hour cadence. While the sync runs on a default schedule, the specific timing can be customized upon request. Specify the desired sync schedule or offsets when submitting the initial Support ticket.