Metrics & Analytics

Understand and track key performance metrics for your data pipelines

Core Metrics
Essential metrics tracked by SEER

Success Rate

Percentage of successful pipeline runs over a given time period. Calculated as (successful runs / total runs) × 100

Execution Time

Average, minimum, and maximum duration of pipeline execution. Helps identify performance degradation

Error Rate

Frequency of failures across all pipelines. Track by error type for targeted improvements

Data Volume

Records processed, rows inserted/updated, or data size transferred during pipeline execution

Custom Metrics
Track application-specific metrics

Adding Custom Metrics

Include custom metrics in the metadata field:

{
  "pipeline_name": "ETL Pipeline",
  "status": "success",
  "metadata": {
    "rows_processed": 10000,
    "new_records": 250,
    "updated_records": 150,
    "execution_time_seconds": 45,
    "memory_used_mb": 512,
    "custom_metric_name": 123
  }
}

Pro Tip: Use consistent metric names across runs to enable trend analysis and alerting

Time-Based Analysis
Monitor metrics over time

Trend Detection

SEER automatically tracks trends for:

  • Execution duration changes over time
  • Success rate patterns (daily, weekly, monthly)
  • Data volume growth or decline
  • Peak usage times and scheduling patterns

Historical Comparisons

Compare current performance against historical baselines to identify anomalies or performance degradation early

Metric-Based Alerts
Set up alerts based on metric thresholds

Alert Types

  • Threshold alerts: Trigger when metric exceeds or falls below a value
  • Trend alerts: Notify on significant changes from baseline
  • Anomaly detection: Alert on unusual patterns
  • SLA violations: Track against service level agreements

Example Alert Rules

  • • Alert if execution time > 10 minutes
  • • Alert if success rate < 95% over 24 hours
  • • Alert if no runs detected in 1 hour (for scheduled jobs)
Dashboard Visualizations
View metrics in the SEER dashboard

Available Charts

  • Success/failure rate over time (line chart)
  • Execution duration trends (area chart)
  • Pipeline status distribution (pie chart)
  • Custom metric timeseries (line chart)
  • Error frequency by type (bar chart)

Tip: Export metrics data as CSV for deeper analysis in external tools like Excel or Tableau