Correlations and Watchdog for Infrastructure
summary
Datadog’s Watchdog automatically models weeks of service metrics, sets bounds of normal, and flags anomalies with no configuration, now extending from APM to infrastructure (beta) to detect memory leaks, TCP retransmit jumps, and integration issues in Redis, Postgres, Nginx, ELB, and S3. A new Correlations tool scans thousands of APM, dashboard, integration, and custom metrics from Watchdog/dashboards/notebooks/monitors, and a demo shows a Lambda latency spike linked to an SQS message surge (tagged ASV network flow automation) and correlated Lambda invocations from a payment management app.
AI-written summary · about
related talks
38mAWS Observability as Code: Leveraging Datadog
24mFrom firefighting to foresight: Powering innovation
33m100% Visibility
12mUnlocking the Secrets of Client Data
46mObservability in the real world: Improve systems w/iterative approach
14mMonitoring Network Traffic in Kubernetes
28mPatterns for scalable security detections on AWS
45mFinding the signal in the noise when running cloud-native apps