Models Dashboard
Last updated
Last updated
The Models Dashboard, part of Paradime's Radar suite, offers comprehensive insights into your dbt™ model performance and dependencies. This tool enables teams to effectively monitor, optimize, and make data-driven decisions to enhance their data pipeline's efficiency and reliability.
Prerequisites
Completed dbt™ Monitoring setup in Radar's Get Started guide.
The Models dashboard is divided into two main sections:
Overview: Provides a high-level summary of all models. [link]
Detailed: Offers in-depth analytics for individual models. [link]
The Overview section gives you a broad perspective on your dbt™ models' performance, allowing you to uncover key insights, including:
Value: Quickly identify time-intensive models and assess overall reliability to find efficiency gains.
How to use:
Monitor the daily success, error, and skipped rates of your dbt™ model runs.
Investigate models with consistently long execution times for optimization opportunities.
Identify patterns in model execution failures to improve overall reliability.
Value: Understand the distribution of materialization types across your models to optimize resource usage and reduce costs.
How to use:
Review the distribution of materialization types (table, view, incremental, snapshot).
Identify opportunities to optimize materialization strategies for better performance and cost-efficiency.
Consider adjusting materialization types for models based on their usage patterns and update frequency.
Value: Pinpoint models that fail frequently to bolster reliability.
How to use:
Focus on models with the highest failure counts for immediate troubleshooting.
Investigate common causes of failures across problematic models.
Implement error handling and notification systems for critical models.
Value: Understand how models are utilized across different schedules.
How to use:
Identify models that are used in multiple schedules, indicating their importance in the data pipeline.
Ensure critical models have appropriate testing and monitoring in place.
Optimize schedules based on model dependencies and execution patterns.
The Detailed section allows you to dive deep into individual model performance, providing comprehensive insights such as:
Value: Analyze the consistency, reliability, and performance of individual models.
How to use:
Monitor the overall success rate and invocation counts to gauge model reliability.
Track the duration of each run over time to identify performance trends.
Investigate any patterns in errors or unusually long run times.
Use these metrics to prioritize which models need optimization or troubleshooting.
Value: Understand execution patterns and identify potential bottlenecks.
How to use:
Identify peak execution times for the model throughout the day.
Consider adjusting run schedules to optimize resource utilization.
Investigate factors contributing to longer execution times during specific hours.
Value: Gain insights into the frequency and success of model executions over time.
How to use:
Observe trends in model invocation frequency.
Correlate invocation patterns with business activities or data refresh schedules.
Identify any unexpected spikes or drops in model usage.
Value: Get a comprehensive view of how the model is executed within different schedules.
How to use:
Review the specific commands and configurations used to execute the model in different schedules.
Analyze execution times across different schedules to identify potential optimizations.
Ensure consistent performance across all schedules that use the model.
Locate the "Select date range" and "Select a model" dropdown at the top of the dashboard.
Choose your desired date range and specific model for analysis.
The dashboard will automatically update to reflect your selections, allowing for focused analysis of specific models over your chosen time period.
