Onboarding Junior Maintainer
This document is the fastest way so you can maintain ADI-JAYA safely.
If you want the 3 day sprint version, go to Day 1 - Day 3 Onboarding. For cross-page ambiguous items, use Verification Matrix.
1) 🎯 System objectives (practical version)
ADI-JAYA handles project flow, procurement, approval and payments that are sensitive to status + role.
Biggest risk:
- change entity status without role validation,
- bulk data changes without backfill plan,
- deploy without post-release verification.
2) 🛡️ Rules of engagement (mandatory)
- Don't change production data directly without a ticket + approval.
- Always start from a separate branch.
- All schema/data changes must have an impact analysis, rollback plan, and post-deployment verification.
- If in doubt, write requires verification in Verification Matrix then escalate.
3) 🧭 Where to find X
- Role & permission → role-access-matrix
- Legal status per domain → state-machine
- End-to-end business flow → business-flows, end-to-end-process
- Daily API (compact) → api-reference
- Full API for auditing → api-full-catalog
- Payload examples → api-examples, request-response-catalog
- Fast DB relations → database-viewer
- Full table inventory → database-full-catalog
- Safe data changes → change-data-playbook, change-example-pack
- Deploy & ops → deployment-runbook, troubleshooting
4) 7 day learning map
Day 1: orientation
Read: overview, architecture, domain-model, module-map.
Day 2: API + roles
Read: api-reference, request-response-catalog, role-access-matrix.
Day 3: data & state
Read: database-viewer, database-design-full, state-machine.
Day 4: surgery
Read: setup-local, deployment-runbook, observability-ops.
Day 5: maintenance simulation
Run at least 3 scenarios from common-maintenance-tasks.
Day 6: data change simulation
Follow change-data-playbook for migration/backfill exercises in staging.
Day 7: troubleshooting drill
Run 3 cases from the troubleshooting decision tree.
5) ✅ Definition of done for juniors
You are considered ready to maintain the system if you can:
- reading logs and hooking to the right module,
- explain the impact of schema changes,
- run deploy check without being guided,
- write rollback notes that other teams can execute.
6) 🚫 Anti-patterns to avoid
- Fast fix without test/verification.
- Change business queries directly in production.
- Removed data because it looked duplicate without relationship analysis.
- Considers enum status trivial.
7) 📣 Escalation
Immediately escalate to a senior if:
- touch financial data / final approval,
- need schema hotfix during peak hours,
- rollbacks can potentially impact active transactions,
- root cause has not been found > 60 minutes.