Skip to content

Learning track

This project is built to learn how an enterprise event platform is designed and operated, and to show why each decision is made so the reasoning can be followed and reused (ADR 0023).

Folder Contents
concepts/ One note per concept: what it is, why it exists, what can go wrong, options, how to decide
retros/ One retrospective per phase: system-level decisions and lessons

Learning path

Phase Concepts Notes
0 · Foundation Decision records, shift-left repository hygiene, supply-chain pinning, regulatory mapping, documentation types Retro
1 · Cluster Infrastructure-as-code state (backend, locking, encryption), the bootstrap problem, secret zero, layered provisioning, GitOps reconciliation OpenTofu state · Secret zero · GitOps reconciliation and ownership · Cattle, not pets · Retro
2 · Platform Policy as code, metrics/logs/traces, secrets management, log retention, TLS automation upcoming
3 · Kafka Replication and ISR, KRaft quorum, TLS/SCRAM and ACLs, schema evolution upcoming
4 · Apps Exactly-once processing, deterministic state machines, order books, idempotency, market surveillance upcoming
5 · Tests Testing a platform at every layer, chaos engineering upcoming
6 · Billion run Performance methodology, end-to-end reconciliation upcoming
7 · Resilience RPO/RTO, failure domains, backup versus replication upcoming
8 · Lifecycle Upgrade ordering, irreversible changes, maintenance windows upcoming