Performance Optimization

EchoSphere is designed to execute tests concurrently. Use the tips below to keep runs fast and stable.

Parallel Execution

• EchoSphere runs tests in parallel to reduce total runtime.
• Keep individual tests lightweight; avoid full table scans when possible.
• Partition large validations by date or key ranges.

Query Tuning

• Filter early and select only needed columns.
• Use appropriate clustering/partitioning in Snowflake to improve aggregation and filter performance.
• Consider pre-aggregations for expensive checks and validate the aggregates instead of raw detail.

Managing Resources

• Choose a warehouse size appropriate for your workload.
• Run heavy suites during off-peak hours where possible.
• Use separate warehouses for CI to avoid contention with production workloads.

Flaky Tests

• Avoid nondeterministic functions unless inputs are fixed.
• Control time-based logic by fixing a date window (e.g., yesterday) instead of CURRENT_DATE when appropriate.

Measuring

• Track duration of suites across runs (your CI can record job duration).
• Export failing rows to analyze patterns of slowness in specific checks.