Production observability
Use logs, metrics, and traces to explain production behavior.
Explain an affected user’s request from recorded evidence
Ana’s save takes 1.2 seconds while the overall dashboard looks normal. You need enough evidence to locate her request and distinguish waiting for a database connection from executing a query. Raw URLs and private payloads are not required to make that relationship visible.
Read a structured event, trace the timed operations, and choose bounded metric dimensions. Extend the evidence across queued work, then discuss sampling, collection failure, and retention. The deliverable is an explanation supported by records, not a dashboard screenshot alone.
Parts group related chapters. Each lesson has a chapter.lesson address, such as 4.07. Open a title below, or use Next to follow the reading sequence. Within a lesson, On this page lists its sections.
- 14.01
Use logs, metrics and traces to explain one slow request
Concepts and examples
- 14.02
Find database-pool waiting in slow API requests
Concepts and examples
- 14.03
Ingest and query metrics with bounded cardinality
Concepts and examples
- 14.04