Long-Running Model Bottleneck Detection
warninglatencyUpdated Mar 2, 2026
Identifies models with the longest execution times that create bottlenecks in dbt job runs, enabling targeted optimization of the most impactful models.
Sources
Debug errors | dbt Developer Hubdocs.getdbt.com
How to Debug dbt Model Failuresoneuptime.com
Optimize and troubleshoot dbt models on Databricks - dbt Docsdocs.getdbt.com
Model performance | dbt Developer Hubdocs.getdbt.com
Run visibility | dbt Developer Hubdocs.getdbt.com
Examining our builds | dbt Developer Hubdocs.getdbt.com
Technologies:
How to detect:
Monitor individual model execution durations from dbt run logs or Model Timing dashboards. Flag models in the top 1% of durations or those exceeding expected thresholds (e.g., >5 minutes for tables, >30 seconds for views).
Recommended action:
Review the flagged model's SQL complexity, consider switching to incremental materialization if full-refresh, optimize joins and aggregations, or increase warehouse size for compute-bound models. Use compiled SQL in target/compiled to analyze query plans.