Monitoring with Tecton

Tecton provides comprehensive tools for monitoring your feature engineering pipelines and their data outputs. This page offers an overview of Tecton's monitoring capabilities and links to detailed documentation for each feature.

Materialization

Materialization jobs can be monitored through the Tecton Web UI, SDK, and CLI. Track materialization status, progress, and performance metrics to ensure your features are up-to-date and available.

Online Serving

Online Serving offers comprehensive monitoring and debugging tools for Tecton's online serving infrastructure. Access detailed dashboards for overall feature serving performance, online store metrics, and feature service-specific analytics. Includes troubleshooting guides for common latency and error scenarios.

Data Quality Metrics

Data Quality Metrics provide quick insights into your Feature View pipelines through summary statistics. Use these metrics to validate data quality and ensure your feature pipelines are producing expected values. Available for both Batch and Stream Feature Views.

Data Quality Validation

Data Quality Validation automatically detects data quality issues after Feature View materialization. The system sends alert emails when validation results indicate data fails to meet expectations during a materialization interval. Supports both Batch and Stream Feature Views.

Alerts

Alerts automatically notify you of issues in your feature engineering pipelines and data. Configure alerts for feature freshness, materialization failures, and data quality issues to maintain system reliability.

Metrics API

Metrics API delivers standardized performance monitoring metrics for the Tecton Feature Platform using the OpenMetrics format. Monitor feature service requests, latencies, errors, and system utilization through integration with popular APM systems like DataDog, SignalFX, and New Relic. Build custom dashboards and alerts to track your specific monitoring needs.