COF-C03 Snowflake Performance Tuning: Query Profile, Pruning, and Scaling

About this video

Slow Snowflake queries do not automatically mean you need a larger warehouse. In this video, we break down how to diagnose Snowflake performance problems and choose the right optimization strategy for the SnowPro Core COF-C03 exam. We’ll cover: Snowflake Query History Query Profile Partition pruning Micro-partition scanning SQL predicate design Projection pruning Disk spilling Warehouse sizing Scale up vs. scale out Multi-cluster warehouses Clustering keys Automatic Clustering Warehouse local cache Auto-Suspend trade-offs Search Optimization Service Query Acceleration Service Materialized Views Performance testing Credit and maintenance costs Common COF-C03 optimization mistakes One of the most important concepts is that different symptoms require different fixes. If a query is scanning too many micro-partitions, increasing warehouse size may not address the root cause. If a single query is spilling intermediate data to disk because of memory pressure, scaling up may help. If many concurrent queries are queueing, scaling out may be the more appropriate response. And if you are considering clustering, Search Optimization, Query Acceleration, or materialized views, you should evaluate both the performance benefit and the ongoing maintenance cost. For COF-C03, use an evidence-first approach: Find the query → inspect the profile → identify the dominant constraint → change one variable → retest under comparable conditions → compare performance and cost The goal is not simply to make queries faster. It is to understand why they are slow and apply the smallest justified change. #Snowflake #COFC03 #SnowProCore