Interface MetricTimeSeriesRepository

All Superinterfaces:

org.springframework.data.repository.Repository<MetricRecordEntity,Long>

@Repository
public interface MetricTimeSeriesRepository
extends org.springframework.data.repository.Repository<MetricRecordEntity,Long>

Spring Data JPA repository providing time-series and aggregated metric queries for the chart and anomaly endpoints.

Contains native SQL queries that aggregate MetricRecordEntity records into 30-second buckets for system-wide chart rendering, and JPQL queries for per-service historical data and recent anomaly retrieval. Uses JOIN FETCH on the service relationship where the controller subsequently accesses service.name to prevent lazy-loading exceptions.

Author:

Nazar Montytskyi

Method Summary

Modifier and Type	Method	Description
List<MetricRecordEntity>	findRecentAnomalies(LocalDateTime from, org.springframework.data.domain.Pageable pageable)
List<MetricRecordEntity>	findRecentByService(Long serviceId, LocalDateTime from, MetricRecordEntity.MetricSource source, org.springframework.data.domain.Pageable pageable)
List<Object[]>	findSystemAggregated(LocalDateTime from, int lim)

Method Details

findRecentByService

@Query("SELECT m FROM MetricRecordEntity m\nWHERE m.service.id = :serviceId\n  AND m.recordedAt >= :from\n  AND m.source = :source\nORDER BY m.recordedAt ASC\n")
List<MetricRecordEntity> findRecentByService(@Param("serviceId")
 Long serviceId,
 @Param("from")
 LocalDateTime from,
 @Param("source")
 MetricRecordEntity.MetricSource source,
 org.springframework.data.domain.Pageable pageable)

findRecentAnomalies

@Query("SELECT m FROM MetricRecordEntity m\nJOIN FETCH m.service\nWHERE m.anomaly = true\n  AND m.recordedAt >= :from\nORDER BY m.recordedAt DESC\n")
List<MetricRecordEntity> findRecentAnomalies(@Param("from")
 LocalDateTime from,
 org.springframework.data.domain.Pageable pageable)

findSystemAggregated

@Query(value="SELECT\n    TO_TIMESTAMP(FLOOR(EXTRACT(EPOCH FROM recorded_at) / 30) * 30)::timestamp AS bucket,\n    AVG(response_time_ms)                      AS avg_rt,\n    MAX(response_time_ms)                      AS max_rt,\n    AVG(cpu_usage)                             AS avg_cpu,\n    AVG(heap_used_mb)                          AS avg_heap,\n    COUNT(DISTINCT CASE WHEN status = 'UP'       THEN service_id END) AS up_cnt,\n    COUNT(DISTINCT CASE WHEN status = 'DOWN'     THEN service_id END) AS down_cnt,\n    COUNT(DISTINCT CASE WHEN status = 'DEGRADED' THEN service_id END) AS deg_cnt,\n    COUNT(DISTINCT CASE WHEN anomaly = true      THEN service_id END) AS anomaly_cnt\nFROM metric_records\nWHERE recorded_at >= :from\nGROUP BY FLOOR(EXTRACT(EPOCH FROM recorded_at) / 30)\nORDER BY bucket ASC\nLIMIT :lim\n",
       nativeQuery=true)
List<Object[]> findSystemAggregated(@Param("from")
 LocalDateTime from,
 @Param("lim")
 int lim)