Data Observability

이 페이지는 아직 영어로 제공되지 않습니다. 번역 작업 중입니다.
현재 번역 프로젝트에 대한 질문이나 피드백이 있으신 경우 언제든지 연락주시기 바랍니다.

Data Observability is in Preview.

Data Observability helps data teams detect, resolve, and prevent issues that impact data quality, performance, and cost. It enables teams to monitor anomalies, troubleshoot faster, and maintain trust in the data powering downstream systems.

Lineage graph showing a failed Spark job upstream of a Snowflake table with an alert and four downstream nodes labeled Upstream issue.

Datadog makes this possible by monitoring key signals across your data stack, including metrics, metadata, lineage, and logs. These signals help detect issues early and support reliable, high-quality data.

Key capabilities

With Data Observability, you can:

Detect anomalies in volume, freshness, null rates, and distributions
Analyze lineage to trace data dependencies from source to dashboard
Integrate with pipelines to correlate issues with job runs, data streams, and infrastructure events

Monitor data quality

Lineage graph centered on the quoted_pricing Snowflake table with an alert on a pricing metric and sidebar charts for freshness, row count, and size.

Datadog continuously tracks metrics and metadata, including:

Data metrics such as null count, null percentage, uniqueness, mean, and standard deviation
Metadata such as schema, row count, and freshness

You can configure static thresholds or rely on automatic anomaly detection to identify unexpected changes, including:

Missing or delayed updates
Unexpected row count changes
Outliers in key metrics

Trace lineage and understand impact

Lineage graph tracing data flow from Kafka through a failed Spark job to a Snowflake table with an alert and four downstream nodes labeled Upstream issue.

Data Observability provides end-to-end lineage, helping you:

Visualize dependencies between tables, columns, and dashboards
Identify upstream root causes and assess downstream impact
Debug faster and make safer schema or transformation changes

Correlate with pipeline and infrastructure activity

Lineage graph showing a failed Spark job with a missing S3 path error, plus a side panel with job run stats and duration trends.

Understand how pipeline activity and infrastructure events impact your data. Datadog ingests logs and metadata from pipeline tools and user interactions to provide context for data quality issues, including: