Anomaly Detection on OpenStack Logs using M.L.
summary
OpenStack logs from a three-controller, two-compute demo (about 150 log files per node) feed an ELK stack: Logstash parses logs with grok into JSON, Elasticsearch indexes them, and Kibana’s x-pack machine learning runs unsupervised single-metric jobs on API response-time (mapped to integer) with 15-minute buckets. Synthetic CPU spikes (dd) raise API latency and produce high anomaly scores; alerts can be sent by email or webhook and Kibana highlights severity and timestamps.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 0:40Who are you
- 1:38Problem Statement
- 2:30Anomaly Detection
- 7:37Issues with OpenStack
- 9:55LK Stack
- 19:02Data Type
- 22:17Demo
related talks
43mBeyond Logs: Real-Time Anomaly Detection with Machine Learning
34mML-Bound Anomaly Detection: Experiences Beyond Commodity Observability
35mAnomaly Detection for Cloud Native Storage
30mOh, Bring Back My Logging Log Anomaly Detection with Opni
21mAnomaly Detection for Platform Logs using Embedded Context
27mBeyond Logs: Real Time Anomaly Detection with Machine Learning
26mAnomaly Detection System for the CERN Cloud
30mBetter data quality with ML & visualization