Concepts → Snowflake View

Overview

In 2026.3.0, Incorta introduced Direct Query Integration, enabling you to query live data directly from source systems such as Snowflake and visualize results instantly without traditional data ingestion or modeling. After configuring your cluster, you can query Snowflake data directly via a Snowflake view or result set, and then create an insight on top of it.

A Snowflake view is a runtime business view that queries data directly from the Snowflake account configured in the Cluster Management Console (CMC)t, without requiring an Incorta physical schema table or business view to reference it.

Create a Snowflake view

The following steps show how to create a Snowflake view after you configure your environment.

1. Log in to the Incorta platform as a Schema Manager.
2. On the navigation pane, select Business Schema.
3. Select a schema that you have Edit access to or create a new one.
4. If required, select Edit, and then select + New > Add New View.
5. In the canvas, select Create via Snowflake.
6. In the Snowflake dialog, enter the SQL query, and then select Execute.
7. Select Done, update the view name, and then select Save Changes.

Referenceable objects in the query

In a Snowflake view, you can reference only tables available in the Snowflake account configured in the CMC. You cannot reference Incorta physical or business schema objects. However, you can reference presentation variables in the WHERE clause using the getVar(var_name, "default_value") function.

Supported SQL and recommended queries

• The Snowflake view supports the standard ANSI SQL syntax.
• You must provide an alias for computed columns, such as SUM() and AVG(), in the SQL query. Failing to use an alias will work in the Snowflake editor. However, you will face query execution failure when exploring data or creating insights on top of the Snowflake view.
• When you run SQL queries against your Snowflake data:
  • Use aggregated queries (such as SUM and AVG).
  • Avoid flat queries (such as SELECT *).

Examples

The following is an example of the SQL queries recommended for Snowflake views:

Copy
SELECT
    STATE,
    SUM(SALES) AS TOTAL_SALES,
    SUM(PROFIT) AS TOTAL_PROFIT,
    RANK() OVER (ORDER BY SUM(SALES) DESC) AS SALES_RANK,
    RANK() OVER (ORDER BY SUM(PROFIT) DESC) AS PROFIT_RANK
FROM SALES_FLAT
GROUP BY STATE
ORDER BY SALES_RANK;

The following is an example of referencing a presentation variable using the getVar() function.

Copy
SELECT
    TPCH_SF1.ORDERS.O_ORDERSTATUS as Order_Status,
    TPCH_SF1.ORDERS.O_ORDERPRIORITY as Priority,
    TPCH_SF1.ORDERS.O_ORDERDATE as Date,
    sum(TPCH_SF1.ORDERS.O_TOTALPRICE) as sales
FROM TPCH_SF1.ORDERS
WHERE Order_Status = getVar(p_sts, "P")
GROUP BY 1,2,3

Insights over Snowflake views

After defining a Snowflake view, you can explore its data and create an Incorta insight based on this view. Most of the features of the Incorta insight are available, except for the following:

• Runtime dashboard filters, including drilling down
• Dashboard filters and prompts, except for presentation variables
  Note

  To apply dashboard filters, use presentation variables directly in the SQL query with the following syntax: getVar(var_name, "default_value").

  This syntax works across all variable data types. No explicit type casting is required as non-string values are cast automatically.

Caching direct queries

To balance real-time data needs with system performance and resource efficiency, Incorta supports caching the direct query results. In addition to the global CMC option, you can specify the Cache Validity Duration at the Snowflake view level, which overrides the global setting, providing flexible control over caching behavior for different workloads.

When a user opens an insight created on top of a Snowflake view, the system checks the cache validity.

• If there is a valid cache, the insight displays cached data.
• If the cache has expired or there is no cache for the underlying query, the system automatically retrieves fresh data from the source, ensuring optimal performance without compromising data freshness.

Limitations

• The result of the Snowflake direct query must not exceed 1 million records. The Snowflake view will be saved; however, insights on top of it will fail to render. To change this limit, contact Incorta Support.
• Data lineage is not yet supported for Snowflake views or their columns. It is not functioning properly for insights over these views either.
• You can explore data from a Snowflake view via Analyzer only. Business notebooks do not support Snowflake views.
• Snowflake views are not verified business views. As a result, Incorta Intelligence and third-party BI tools will not have access to it through Advanced SQLi.
• Calculations within the direct query are handled by Snowflake and follow standard SQL null handling. However, calculations (formula columns) added at the insight level are handled by Incorta and respect its null handling behavior based on the Null Handling CMC option.
• You cannot edit or change the view column data type. When you change the data type of the source column, you must save the Snowflake view to have the data type updated.