Observability Fundamentals
Telemetry contracts in code review
Turn observability expectations into reviewable artifacts: diffs, schemas, and CI hints that nudge teams before merge.
- Format
- Cohort
- Duration
- 5 weeks, cohort
- Skill
- Advanced
- Stack
- GitHub Actions, OTel
Tuition (informational): KRW 180,000
Request a syllabus conversation
You write small policy files checked in beside services: required span names for outbound calls, mandatory log fields for errors, and feature flags that gate experimental cardinality. CI exercises fail harmlessly first, then you tighten rules as a pretend org matures. The outcome is review comments that feel kind because the robot complained first.
What is included
- Example CI jobs that lint OpenTelemetry resource attributes
- Pull request templates with observability checklist sections
- Pair exercises writing kind failure messages for policy breaks
- Discussion on ownership when platform and product disagree
- Lightweight schema previews for log fields
- Capstone proposing a policy MR to a sample monorepo
- Retro on when policies should sunset
Outcomes
- You open a policy MR your team would realistically merge.
- You explain a CI observability failure to a product engineer without jargon pile-on.
- You schedule one policy sunset with criteria documented.
Instructor of record
Brings code review scars and a love for robots that nag politely.
Jonah Park
Primary feedback on labs
Participant questions
Examples use a monorepo for speed, but mapping to many-repo orgs is covered in the final week.
Recent voices
“CI nag caught a missing service.name before human reviewers even scrolled to the diff.”