observability-service-health

Diagnose APM service health by querying Observability APIs, ES|QL, and Elasticsearch for SLO status, alerts, ML anomalies, throughput, and error rates. Correlate these metrics with infrastructure and logs to perform root cause analysis on latency spikes.

Investigate SLO violations and firing alerts by synthesizing APM performance metrics with dependency mapping. Utilize the APM annotations API and apm-correlations script to identify if a downstream bottleneck is causing the service to slow down.

Yes, use ES|QL to query Elasticsearch for throughput and error rates during root cause analysis. Combine these queries with ML anomaly detection and log correlation to pinpoint underlying infrastructure issues impacting service health.

Yes, comprehensive service health assessment requires correlating APM data with infrastructure and logs. This correlation is essential to move beyond symptom detection and accurately identify the underlying root cause of performance degradation.

ML anomalies can appear without SLO violations when throughput, latency, or error rates deviate from learned baselines but remain below SLO thresholds. Check the APM annotations API to investigate these early warning signs before they impact SLOs.