Structured Logging

Explore structured logging techniques in this code::dive 2017 conference talk by Bartosz Szurgot. Delve into the evolution of logging practices, from simple string searches to more sophisticated approaches. Examine various logging APIs, data structures, and output formats, including JSON. Learn how to implement a structured logging system that supports basic types, multi-field types, and nested structures. Discover the benefits of maintaining log structure throughout the logging process and how it can enhance system monitoring and debugging. Gain insights into practical implementation strategies, including the use of recursive variadic templates and Docker integration. Understand the broader implications of structured logging in the context of modern logging ecosystems like ELK (Elasticsearch, Logstash, Kibana).

Syllabus

Wprowadzenie
A tale of a string search
String search - sentinel
Just logging?!
Tale of an addressing scheme...
Coding part...
Data structure!
System activity log
stream-like?
printf-like?
Output?
Problems to solve
Expected output
JSON output example
The Difference
Possibilities
Simple log API . Raw arguments
"Formatted" log API
Your fields - your wrapper!
Logs' destination
Logical view
Recursive, variadic template?!
User's perspective
Basic types
User API-simple types
User API-multi-field types
User API-nested types struct Vector 2D
Logger with format
String parsing
Glue it up!
Formatting
Usage
Translations
Logs are coming!
Keep'em structured!
Example implementation
SDK with Docker talk
Universal idea
What's on the menu? (ELK-referenced)