The keycloak project ended with four open questions that design could not
settle. A two-VM lab was built to answer them by measurement, and this is
that material: 26 experiments, 125 raw command outputs, 22 browser captures.
Follows the import procedure in README.md.
source/ the originating repository verbatim — 78 documents, 28 SVGs,
8 manifests, plus .source-revision recording the commit
final/ the SSOT
document.md 729 lines written from the 29 experiment documents, not
concatenated: what was predicted, what was measured, and
where the measurement itself was wrong
evidence/raw 125 outputs, flattened to <experiment>__<file> because
the originals collided (01-baseline.txt appeared three
times) and the audit only globs the top level
evidence/meta one per raw file; command and exitCode are null and the
README says why rather than inventing them
evidence/browser 22 captures
assets/ three diagrams through techviz
.techviz/ their VizSpecs
A separate project rather than an addition to keycloak: the B-layer answers
that project's four questions, but the A, C and D layers are about cluster
failure, SSO and operations, and one document.md should hold one subject.
The four question records there can point here through 관계.
Recorded rather than papered over: only three of the 28 diagrams were
remade. The repository forbids hand-drawn SVG and forbids titles inside the
canvas; all 28 originals carry both, so converting them is redrawing, not
reformatting. They stay in source/ and the gap is written into the document.
verify-pipeline.py passes. audit-records.py reports no issues.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
54 lines
3.1 KiB
Markdown
54 lines
3.1 KiB
Markdown
---
|
|
kind: CONCEPT
|
|
slug: messaging-outbox-jdbc-postgresql-c01
|
|
title: 모르는 것을 실패로 취급하지 않는다
|
|
topic: transaction-and-consistency-models
|
|
project: clean-architecture-backend-template
|
|
status: 게시 전
|
|
sourceRevision: 21234e38cdb9a926cbc92bb97a2aee2e4a7d2916
|
|
rootTreeNode: concept:messaging-outbox-jdbc-postgresql-c01
|
|
evidenceCapturedOn: 2026-09-01
|
|
assets:
|
|
- key: messaging-outbox-jdbc-postgresql-c01
|
|
file: ../../../final/evidence/rendered/messaging-outbox-jdbc-postgresql-c01.svg
|
|
- key: messaging-outbox-jdbc-postgresql-c01-diagram
|
|
file: ../../../final/assets/diagrams/messaging-outbox-jdbc-postgresql-c01.svg
|
|
evidence:
|
|
- ../../../final/evidence/raw/messaging-outbox-jdbc-postgresql-c01.txt
|
|
source:
|
|
- 원본 분석 절은 analysis/messaging/messaging-outbox-jdbc-postgresql.md#L66 이다.
|
|
module: messaging-outbox-jdbc-postgresql
|
|
---
|
|
|
|
# 모르는 것을 실패로 취급하지 않는다
|
|
|
|
트랜잭셔널 아웃박스의 PostgreSQL 구현이다. 이 리프의 축은 하나다 — 모호한 발행을 같은 메시지 id로 재시도한다.
|
|
|
|
## 본문
|
|
|
|
<!-- body:start -->
|
|
|
|
**트랜잭셔널 아웃박스의 PostgreSQL 구현**이다. 비즈니스 트랜잭션이 쓰고 릴레이가 배출한다. 이 리프의 축은 하나다: **"모르는 것을 실패로 취급하지 않는다."**
|
|
|
|
> "The relay's correctness rests on one rule: an ambiguous publish is retried **under the same message id**. Minting a new id would turn a possibly-delivered message into a definitely-second message, and no downstream deduplication could recover from it. Marking it failed instead would lose a message the broker may already hold."
|
|
|
|
마지막 문장이 중요하다 — 릴레이는 at-least-once publication만 보장한다고 자기 상한을 스스로 명시한다.
|
|
|
|
## 이 리프가 아는 것과 모르는 것
|
|
|
|
:::evidence key="messaging-outbox-jdbc-postgresql-c01-diagram" alt="리프 경계 안에 발행 포트와 관계형 데이터베이스가 들어 있고 브로커와 스프링 컨텍스트가 경계 밖 점선 상자로 놓인 구조" caption="이 리프가 아는 것과 모르는 것" zoom="false"
|
|
:::
|
|
|
|
경계: 브로커를 모른다(`MessagePublisher` 포트만 안다). 스프링 컨텍스트를 모른다(`spring-jdbc`/`spring-tx` 는 `implementation` 이며 트랜잭션 동기화 조회에만 쓴다). 배선은 starter 몫이다.
|
|
|
|
## MessagePublisher 참조 위치
|
|
|
|
:::evidence key="messaging-outbox-jdbc-postgresql-c01" alt="코드베이스에서 MessagePublisher 를 검색한 출력 14줄. 이 기록이 세는 참조가 그 출력에 그대로 보인다." caption="MessagePublisher 코드베이스 검색 — 14줄 · exit 0" zoom="true"
|
|
:::
|
|
|
|
## 파괴적 작업 저널이 같이 사는 이유
|
|
|
|
`messaging-admin-api` 의 파괴적 작업 저널 구현도 이 리프에 산다. build.gradle 이 그 이유를 적는다 — "The destructive-operation journal lives here because it needs exactly what the outbox needs: one relational database every replica can see, and a migration lane that already exists. The contract it implements belongs to the admin API."
|
|
|
|
<!-- body:end -->
|