Snowflake Cost Optimization
Understanding Snowflake Pricing
Snowflake charges for compute (credits based on warehouse size and runtime) and storage ($23/TB/month for on-demand). Compute is typically 80% of the bill. A medium warehouse (4 credits/hour) running 8 hours/day costs ~$1,500/month. Most overspending comes from oversized warehouses and long-running queries.
Warehouse Sizing Strategy
Start small (XS or S) and scale up only if query performance is unacceptable. Use multi-cluster warehouses for concurrent workloads instead of oversizing single warehouses. Set auto-suspend to 60 seconds (not the default 600). Create separate warehouses for different workloads (ETL, BI, ad-hoc).
Query Optimization
Use Snowflake's Query Profile to identify bottlenecks. Avoid SELECT * — columnar storage means fewer columns = less cost. Use LIMIT for exploratory queries. Leverage result caching (no-cost re-runs within 24 hours). Use clustering keys for large tables (>1TB) to improve pruning.
Resource Monitors
Set resource monitors at account and warehouse levels. Configure alerts at 75% and hard stops at 100% of monthly credit budgets. Monitor per-user and per-query credit consumption. Audit warehouse usage weekly to find idle warehouses.
How Cloptima Will Help
Cloptima's Snowflake cost optimization (coming soon) will analyze query patterns, identify expensive warehouses, recommend right-sized configurations, and provide AI-powered query suggestions — the same approach we use for BigQuery optimization today.