Lightning Talk: A Drop-in System To Accelerate Metrics Observability by 100x Us...
summary
Claims a 100x reduction in query cost and latency by computing streaming precomputation and sketch-based approximations as a drop-in accelerator for an existing Prometheus–Grafana stack using Prometheus remote_write and a passthrough fallback to Prometheus for unsupported queries. Uses compact summaries such as a KLL sketch summary that can compress 100 million float samples (~800 MB) to under 10 KB with ~99% quantile accuracy. Prototype runs via Docker Compose; open-source release expected end of 2025.
AI-written summary · about
chapters (9)
- 0:00ASAP query introduction
- 1:05ASAP query integration
- 2:31100x acceleration secret
- 4:57ASAP query deployment
- 6:45Live demo of ASAP query
- 8:15Try ASAP query yourself
- 8:56Final thoughts on visibility
- 9:18Q&A: Accuracy trade-offs
- 10:42Q&A: Cost and performance
related talks
12mLightning Talk: A Drop-in System To Accelerate Metrics Observability by 100x Us...
09mGraduated Project Lightning Talk: Using Trace Data for Monitoring & Alerting of Appl...
11mLightning Talk: Beyond Billions: Operating Thanos, Prometheus & OpenTelemetry...
36mObservability with Prometheus and Beyond: Leveraging Cloud-native Technologies Outside of Cloud
02mLightning Talk: Prometheus adoption stats
05mLightning Talk: Improved Prometheus filtering in Grafana
23mAutomatically Making Dashboards Load 100X Faster
13mLightning Talk: Observability-First DevSecOps: Building Resilient Multi-Cloud P...