Insights Direct Connect FAQs
The questions below relate to Insights Direct Connect for the areas:
- Usage Limits and Compute Hours
- Unified Data Model
- Data Safety and Permissions
- Usage of Data and Integration Options
- Backend Architecture
Usage Limits and Compute Hours
Q. When does compute time begin counting?
Compute time begins when Snowflake executes a query. Connecting alone does not consume compute.
Q. Does compute start when refreshing data?
Yes. Any refresh that triggers Snowflake to run queries consumes compute.
Q. Does navigating or previewing tables consume compute?
Yes. Previewing tables or expanding fields triggers metadata queries that consume compute.
Q. Do background or automatic queries consume compute?
Yes. Any query executed by BI tools counts toward compute usage.
Q. Is compute allocation per user, per tenant, or shared?
Compute is shared at the warehouse level across all users.
Q. Are compute hours consumed differently for DirectQuery vs Import?
DirectQuery consumes compute continuously. Import consumes compute only during refresh.
Q. Are compute hours consumed during background refreshes?
Yes. Any refresh operation consumes compute.
Q. Is there a way to develop dashboards without consuming compute?
No. Any action that triggers a Snowflake query consumes compute.
Q. Best practice for managing compute during development?
Use Import mode, disable auto‑refresh, and build in Power BI Desktop.
Q. What actions consume compute?
Any Snowflake query including previews, refreshes, metadata checks, and DirectQuery interactions.
Q. Do auto‑refreshes count as billable compute?
Yes.
Q.Does previewing tables consume compute?
Yes.
Q. Is there a hard stop when compute limits are exceeded?
Snowflake suspends the warehouse when compute is exhausted.
Q.How can usage be monitored?
Through Snowflake’s usage dashboards and query history.
Q.Is there a usage log?
Yes. Snowflake provides logs showing users, queries, and compute consumption.
Q.How to temporarily disable Direct Connect?
Use Import mode, disable auto‑refresh, or close the Snowflake connection.
Unified Data Model
Q. Can you provide the Unified Data Model?
Yes, see the articles Insights Data Moduel Overview, Insights Unified Data Model and the Insights Unified Data Model - History Tables.
Q. Support for table interpretation?
Documentation is provided.
Q. Support for schema understanding?
Schema documentation is included.
Q. Support for data lineage?
High‑level lineage is documented.
Q. Support for performance optimization?
BI optimization is not included; customers manage performance.
Data Safety and Permissions
Q. Is the Direct Connect is read‑only?
Yes. It is strictly read‑only.
Q. Does Direct Connect respect application permissions?
It provides access to the Insights dataset; it does not inherit user‑level permissions.
Q. Can tables be restricted by role?
Not at this time.
Usage of Data and Integration Options
Q. Can Microsoft Excel connect via ODBC?
Yes, see Connect using Insights Direct Connect from Microsoft Excel to Snowflake.
Q. Does Excel usage consume compute?
Yes.
Q. Can other BI tools connect?
Yes. Any tool supporting Snowflake ODBC/JDBC can connect.
Q. Restrictions for non‑Power BI tools?
No restrictions from Eptura.
Q. How frequently is data updated?
Daily.
Q. How long is data retained?
Retention aligns with Insights policies.
Q. Are historical tables truncated?
Yes. Some have defined retention windows.
Q. Are there query limits or throttling?
Snowflake may throttle or suspend queries.
Q. Is ODBC traffic encrypted?
Yes.
Q. Can conditional access or IP restrictions be used?
Yes.
Backend Architecture
Q. Can you clarify the backend accessibility?
Direct Connect exposes a curated analytical dataset separate from operational systems.
Q. Does API and Direct Connect use the same backend?
No. API uses operational data; Direct Connect uses the Insights analytical dataset.
Q. Why not expose via OData?
OData is not suited for analytical datasets or high‑volume BI workloads.
Q. Why was the API request declined?
A combination of technical constraints and product strategy.
