Metric Monitoring dashboard
Last updated
Was this helpful?
Last updated
Was this helpful?
As soon as a data source is connected, the metric monitoring dashboard is available and will have historical information. Soda establishes a statistical baseline for each metric and continually compares new scan results against that baseline, flagging anomalies according to the sensitivity, exclusions, and threshold strategy you’ve configured.
Soda offers two main types of monitors to support scalable, layered observability: dataset monitors and column monitors.
Together, they offer broad coverage and targeted insight, helping teams detect both systemic and localized data quality issues.
Each of these sections will contain summarized information about the latest scan results for each monitor. From this health tab, you can access each monitor for further investigation and configuration, as well as creating alerts.
You can turn any metric monitor into a proactive alert by clicking its bell icon on the Metric Monitors dashboard and selecting Add Notification Rule. This brings up the Add Notification Rule panel:
Name Enter a descriptive title for your rule (e.g. “Row-Count Alerts – Prod Sales”).
Data source Choose the warehouse or connection to scope your rule. Then, search for and check the specific tables (or columns) this rule should cover. The “Matches X datasets” badge updates in real time so you know exactly what you’ll be alerting on.
Applies to Pick which check type you want to alert on.
Recipients Select one or more notification targets:
Email addresses
Slack channels
Other integrations
This dialog lets you reuse a single rule for multiple datasets or checks, ensuring your team only gets the notifications they care about.
Metric monitors are the foundation of data observability in Soda. Monitors are time-series records of dataset-level metrics that maintain historical values for analysis. Soda automatically collects these metrics and tracks how they evolve over time through a to identify when metrics deviate from expected patterns. These deviations are surfaced in the Metric Monitors tab for each dataset.
provide instant, no-setup monitoring based on metadata. They track high-level metrics like row count changes, schema updates, and insert activity, making them ideal for catching structural or pipeline-level issues across large numbers of datasets.
are more granular and customizable. They focus on specific fields, allowing users to monitor things like missing values, averages, or freshness. These monitors are useful for capturing data issues that impact accuracy or business logic at the column level.
