Snowflake Performance
Snowflake Performance gives you real-time visibility into query performance, resource usage, and system health, so you can find and fix inefficiencies before they affect cost or reliability.
Snowflake Performance has four tabs: Tables, Warehouse Insights, Data Loading, and Query and Task Failures.
Filters
The Data Source Filter switches between connected Snowflake accounts. Organization Unit and Cost Center filters are also available to refine the graphs, dashboards, and tables on this page.
Tables
Shows the 50 most-accessed tables in your Snowflake account for the selected time range, both as a bar chart and a detail table.
Widget | Description |
|---|---|
Table Name | The table's name. |
Rows / Columns | Row and column counts. |
Query Count | Queries executed against the table. |
Average Partitions Scanned | Should be well below the table's total available partitions — a high ratio suggests inefficient scanning. |
Average Query Latency | Average latency experienced by queries on this table. |
Bytes Scanned | Total data scanned — a proxy for query efficiency. |
Clustering Key | The key(s) used to organize the table's data for query performance. |
Top Filtered Key | The most common key used in query filters against this table. |
Select a table name to open its Details page, showing metadata and clustering information that can inform optimization decisions — for example, which columns are accessed most, and whether the current clustering key still fits that access pattern.
Warehouse Insights
Warehouse performance data. Apply the Global Calendar for a specific time range, plus Warehouse and Database filters. If all warehouses are selected in the Filters section, individual charts let you narrow further; if specific warehouses are already selected there, chart-level filtering is disabled and every chart reflects only those warehouses.
Widget | Description |
|---|---|
Total Queries | Total queries executed and total users who ran them, for the selected period. |
Query Execution Times | Average, 90th percentile, and maximum execution time, plus average query size in bytes and average execution time per byte. |
Cost for Selected Time Period | Compute cost and cloud service cost for the period. |
Query Provisioning Wait Times | Number of delayed queries, average/90th percentile/max wait time, and the percentage of queries affected by warehouse provisioning delays. |
Query Overload Wait Times | The same wait-time breakdown, for delays caused by warehouse overload instead of provisioning. |
Query Count By Query Type | Query count by type (SHOW, SELECT, UNLOAD, GRANT, and others). |
Query Cost By Query Type | Cost by query type — helps identify the most resource-intensive query types. |
Average Bytes Per Query By Warehouse Size | Average query size by warehouse size (X-Small, Small, Medium, and so on). |
Average Duration Per Query By Warehouse Size | Average execution duration by warehouse size. |
Top 15 Users | The 15 users with the most executed queries. |
Top 15 Frequent Queries | Fingerprint, Query Count, and partial Query Text for the most-repeated queries. |
Warehouse Metrics | Per-warehouse detail: Warehouse Name, Total Query Count, Total Queued Queries, Warehouse Size, Execution Time, Compute Credits, Cloud Service Credits, Queue Provisioning Time, and Queue Overload Time. |
Credits Used | Snowflake credits consumed, by warehouse, as a pie chart. |
Warehouse Load | Percentage of time each warehouse spent Suspended, at Less Load, or at High Load. |
Top 10 Queries with Spillage | The 10 queries that spilled the most data to disk, for the selected time range. |
Data Loading
How data is loaded into your Snowflake account. Apply the Global Calendar, plus Database, Schema, or Table filters.
Data Loading table:
Column | Description |
|---|---|
Date | When the data was loaded. |
Database / Schema / Table | Where the data was loaded to. |
Ingest Method | The method used to ingest the data. |
Row Count / Rows Parsed | Rows added and rows parsed during ingestion. |
Average File Size / Total File Size | File size statistics for the ingested files. |
Copy Into Tables table:
Column | Description |
|---|---|
Table | The target table. |
Copy Count | Number of COPY INTO executions on the table. |
Execution Time | Total time to complete the operation. |
Bytes Scanned | Data scanned during the operation. |
Rate (MBs) | Data transfer rate. |
Rows Produced | Rows produced by the operation. |
Query and Task Failures
Errors across Snowflake operations. Apply the Global Calendar to scope the data.
Widget | Description |
|---|---|
Failed Queries | Error Code, Error Message, Count, and Total Time for failed queries. Filterable by error code. |
Failed Tasks | Details of failed Snowflake tasks. |
Long Running Tasks | Tasks that take significantly longer than expected to complete. |
Stale Streams | Streams referencing outdated data that no longer provide accurate change tracking. |
What's next
Snowflake Query Studio – Drill into the specific queries behind these performance metrics.
Snowflake Warehouse – Act on warehouse sizing based on the load patterns here.

Send a comment