Centralized Logging Patterns
summary
Five concrete patterns for centralizing logs with Elastic Stack: parsing raw files via Filebeat → Logstash → Elasticsearch → Kibana (fixes multi-line stack traces with a multi-line setting and explains why a broken stack created 42 vs 40 entries), sending logs over the network from logback (risk of dropped lines when Logstash is down), producing structured JSON logs to avoid parsing, running Filebeat as a Docker sidecar with Docker socket for metadata, and deploying Filebeat as a Kubernetes daemonset.
AI-written summary · about
chapters (10)
- 0:00Introduction to centralized logging
- 2:48ELK stack overview and tools
- 5:14Anti-patterns in logging
- 5:57Parsing logs with Logstash
- 12:45Sending logs over network
- 15:34Structured logs in JSON
- 17:48Logging in Docker containers
- 23:26Logging in Kubernetes
- 25:33Summary of five patterns
- 26:17Q&A session
related talks
47mCentralized Logging Patterns
49mCentralized Logging Patterns
51mCentralized Logging Patterns
40mCentralized Logging Patterns
44mCentralized Logging Patterns
14mCentralized Logs using ELK (Elasticsearch, Logstash, Kibana) Stack
42mMake your logs work for you
43mPractical Application Logging and Monitoring