Skip to main content

Performance

The Performance page gives you a complete view of how your Snowflake workload behaves over time - execution speed, efficiency, data processing patterns, and workload trends.

1. Query Volume & Performance​

Focuses on the speed of your Snowflake workload.

Average Execution Time

  • Track average daily execution time vs. a rolling 30-day baseline
  • Identify slowdowns, spikes, or trends
  • Validate improvements after query changes or warehouse resizing

Max Execution Time

  • Identify inefficient logic, large scans, or bad joins
  • Compare max execution time to 30-day norms
  • Spot regressions after ETL or dbt changes

Each data point is clickable - opens a drill-down with the top queries that influenced execution time that day.

Query volume and performance

2. Resource Efficiency & Data Processing​

Explains how efficiently your data is being processed, which directly influences cost and performance.

GB Scanned Over Time - daily total GB scanned across all queries, overlaid with 30-day average.

GB Spilled to Remote Storage - total GB spilled per day vs. historical norms.

Number of Queries Spilled - helps determine whether spilling is isolated or systemic.

Resource efficiency and data processing

Measures how efficiently Snowflake warehouses convert credits into work.

Execution Minutes Per Credit - higher = more efficient.

Number of Queries - detect usage spikes or drops and correlate workload changes with performance changes.

Workload efficiency trends

Drill-Downs & Investigations​

Most performance views let you click into a specific day to investigate what drove the trend. Inside the drill-down:

  • Top 10 queries contributing to that metric
  • Exact user, warehouse, and usage details for each query
  • Open the exact query in Snowflake to investigate further

Performance drill-down

Date Controls & Warehouse Filtering​

  • Choose any date range to explore trends over time
  • Filter by individual warehouses to isolate their performance
  • Highlight only Yuki-optimized warehouses to understand optimization impact

How to Use This Page​

Detect a slowdown: Look for spikes in execution time → click to reveal heavy queries → open in Snowflake.

Validate optimization impact: Compare execution time or minutes per credit before/after enabling Yuki.

Diagnose cost inefficiency: Analyze GB scanned trends and correlate spikes with job/ETL/dbt activity.