Live, Laugh, Log
summary
Operational guide to Fluent Bit with concrete knobs, failure modes and diagnostics: how CRI log files hit the tail input, why the Kubernetes filter enriches records, default max_size = 32 KB and how skip_long_lines prevents stalls, chunk/file-system buffering (up/down/busy) to avoid drops, throttle filter and retry caps for delivery, multi-threading and processors for throughput, plus dry-run, rubular, expect plugin, Prometheus metrics, dump_internal and tap traces for debugging and what a 250% log growth implies for retention and cost.
AI-written summary · about
chapters (18)
- 0:00Welcome to Live Laugh Log
- 1:40Chaotic verbose logs
- 3:27Why logs are essential
- 7:48Introducing Fluent Bit
- 8:51Fluent Bit features overview
- 9:53What is Fluent Bit
- 11:15Deploying in Kubernetes
- 12:40Cloud-native log lifecycle
- 18:42Four log problems to avoid
- 21:50Reliability and performance
- 23:14Validating configurations
- 25:44Buffering for optimal ingestion
- 30:16Throttle filter for log flow
- 32:04Multi-threading boost
- 34:34Monitoring with meta telemetry
- 37:42Diagnosing issues
- 39:05Fluent Bit future and resources
- 40:09Closing remarks and Q&A
related talks
21mWhy Do My Logs Suck?! A Practical Guide to Improving Your Logs
22mLogging: Fluentd & Fluent Bit
36mDeep Dive: Fluent Bit
39mMaking Logs Work for you with Fluentd
34mFluentd: A Complete Logging Ecosystem for Kubernetes
27mUnder the Hood with Fluent Bit Operator: Kubernetes-native Log Processor
35mIntro: Fluentd
21mLogging Operator the Cloud Native Fluent Ecosystem