Docs are not available for mobile use. Please use a desktop computer to view the documentation.

Boundary Detection Runs

Batch processing metadata for boundary detection operations

Purpose

The boundary_detection_runs table tracks the execution and performance of boundary detection batch processes. It provides metadata about detection runs, including timing, processing statistics, and execution status for monitoring and optimization purposes.

Schema Definition

Field	Type	Description	Constraints
`id`	uuid	Unique run identifier	Primary key, auto-generated
`userId`	uuid	User whose data was processed	Foreign key to users.id
`runDate`	timestamptz	Date/time of the detection run	Not null
`status`	detection_status	Current run status	Default: ‘pending’
`totalSignalsProcessed`	integer	Number of signals analyzed	Optional
`boundariesDetected`	integer	Number of boundaries found	Optional
`processingTimeMs`	integer	Total processing time in milliseconds	Optional
`startTime`	timestamptz	When processing began	Optional
`endTime`	timestamptz	When processing completed	Optional
`createdAt`	timestamptz	Record creation timestamp	Auto-generated

Relationships

References:

users.id via userId - User whose data was processed

Enumeration Values

Detection Status

pending - Run is scheduled but not started
completed - Run finished successfully
failed - Run encountered errors and failed

Usage Examples

Get recent detection runs for a user

SELECT runDate, status, totalSignalsProcessed, boundariesDetected, processingTimeMs
FROM boundary_detection_runs 
WHERE userId = 'user-uuid'
ORDER BY runDate DESC
LIMIT 10;

Monitor run performance

SELECT 
  DATE_TRUNC('day', runDate) as day,
  COUNT(*) as total_runs,
  COUNT(*) FILTER (WHERE status = 'completed') as successful_runs,
  COUNT(*) FILTER (WHERE status = 'failed') as failed_runs,
  AVG(processingTimeMs) FILTER (WHERE status = 'completed') as avg_processing_time_ms
FROM boundary_detection_runs 
WHERE userId = 'user-uuid'
  AND runDate >= NOW() - INTERVAL '30 days'
GROUP BY day
ORDER BY day;

Find long-running processes

SELECT 
  id,
  runDate,
  status,
  startTime,
  endTime,
  EXTRACT(EPOCH FROM (COALESCE(endTime, NOW()) - startTime))*1000 as actual_duration_ms,
  processingTimeMs
FROM boundary_detection_runs 
WHERE userId = 'user-uuid'
  AND startTime IS NOT NULL
  AND (
    status = 'pending' OR 
    EXTRACT(EPOCH FROM (COALESCE(endTime, NOW()) - startTime))*1000 > 300000 -- > 5 minutes
  )
ORDER BY startTime DESC;

Analyze detection efficiency

SELECT 
  status,
  AVG(totalSignalsProcessed) as avg_signals,
  AVG(boundariesDetected) as avg_boundaries,
  AVG(processingTimeMs) as avg_time_ms,
  AVG(CASE 
    WHEN totalSignalsProcessed > 0 
    THEN processingTimeMs::float / totalSignalsProcessed 
    ELSE NULL 
  END) as avg_ms_per_signal
FROM boundary_detection_runs 
WHERE userId = 'user-uuid'
  AND status = 'completed'
  AND runDate >= NOW() - INTERVAL '30 days'
GROUP BY status;

Find failed runs for troubleshooting

SELECT runDate, startTime, endTime, totalSignalsProcessed, processingTimeMs
FROM boundary_detection_runs 
WHERE userId = 'user-uuid'
  AND status = 'failed'
ORDER BY runDate DESC;

Get processing statistics

SELECT 
  COUNT(*) as total_runs,
  SUM(totalSignalsProcessed) as total_signals,
  SUM(boundariesDetected) as total_boundaries,
  AVG(boundariesDetected::float / NULLIF(totalSignalsProcessed, 0)) as avg_boundaries_per_signal
FROM boundary_detection_runs 
WHERE userId = 'user-uuid'
  AND status = 'completed'
  AND runDate >= NOW() - INTERVAL '30 days';

Monitor current processing

SELECT 
  id,
  runDate,
  startTime,
  EXTRACT(EPOCH FROM (NOW() - startTime))*1000 as elapsed_ms,
  totalSignalsProcessed
FROM boundary_detection_runs 
WHERE status = 'pending'
  AND startTime IS NOT NULL
ORDER BY startTime;