Observability in a microservice world

Transcript

2. 3 Amigos
3. “Where is my order?“ “Why is the cart slow?“
5. What's the deal? Where is the bu?
6. Drawback of Resilience
7. Known Unknowns v. Unknown Unknowns
8. O11y to the rescue
9. Observe ability
11. Building Blocks Centralized Logging Metrics Tacing
13. Goal be epared to answer the why
18. Collecting and storing data over a period of time
19. Readers Perspective
20. Log Level log.debug -> log.forTestPurpose log.info -> log.asExpectedByDefault log.warn -> log.toInvestigateTomorrow log.error -> log.wakeMeUpInTheMiddleOfTheNight
21. Log Rotation
22. Unstructured Log
23. Structured Log
24. Mapped Diagnostic Context MDC.put("CartEntry", cartEntry.toString());
25. Debugging
26. Insight for Non-Dev
27. Container
28. Centralized Data Correlation
30. Inspect and Adapt xisting Knowledge Configuration Adaptability Dependency OpenSource
43. Data obtained from measuring events Response Time Request Rate
44. Explicit instrumentation
45. Tools Actuator natural choice for Spring Boot apps Since 2.0 based on Micrometer.io Backport available for Spring Boot <2.0
46. Counter
47. Gauge
48. Timer
49. Usage
50. Transport Application metrics need to be exported
51. Dimensional vs Non-dimensional http_server_requests_seconds_count{status="200"} http.server.requests.status.200
52. Fallacies with probabilities 0.99100 ≈ 0.37
62. Request scoped information
63. Correlation via unique IDs
64. Spans and Traces
65. Span Propagation
66. Spring cloud sleuth Tracing instrumentation for Spring Boot apps
67. What you get Span/Trace-Id generation Log integration API for managing custom spans Zipkin integration