[G1-Sync] Manual knowledge update

This commit is contained in:
Antigravity Agent
2026-05-10 22:08:15 +09:00
parent 21ac3ed255
commit 504fd5fb42
3011 changed files with 380280 additions and 206977 deletions
@@ -2,85 +2,273 @@
id: wiki-2026-0508-enterprise-software-engineering
title: Enterprise Software Engineering
category: 10_Wiki/Topics
status: needs_review
status: verified
canonical_id: self
aliases: [P-Reinforce-AI-SOFTWARE-ENG]
aliases: [enterprise SE, SDLC, large-scale software, enterprise architecture]
duplicate_of: none
source_trust_level: A
confidence_score: 0.98
tags: [SoftwareEngineering, Enterprise, SDLC, Process]
confidence_score: 0.95
verification_status: applied
tags: [software-engineering, enterprise, sdlc, process, architecture, scale]
raw_sources: []
last_reinforced: 2026-04-20
last_reinforced: 2026-05-10
github_commit: pending
inferred_by: Claude Opus 4.7 (auto-normalize 2026-05-08)
tech_stack:
language: unspecified
framework: unspecified
language: Universal
applicable_to: [Enterprise, Large-scale, Regulated]
---
# [[Enterprise-Software-Engineering|Enterprise-Software-Engineering]] (엔터프라이즈 소프트웨어 공학)
# Enterprise Software Engineering
## 📌 한 줄 통찰 (The Karpathy Summary)
> "코딩(Coding)이 아닌 엔지니어링(Engineering)의 영역." 상업적 가치를 지닌 소프트웨어를 정해진 예산과 기한 내에 일률적인 품질로 생산하고 유지보수하기 위한 학제적 접근이다.
## 한 줄
> **"매 scale + 매 compliance + 매 long-lifecycle 의 software"**. 매 startup velocity 의 X — 매 audit + 매 SLA + 매 multi-team. 매 modern: 매 platform engineering + DevSecOps + observability + AI-augmented (Copilot, Cursor, Anthropic).
## 📖 구조화된 지식 (Synthesized Content)
- **SDLC (Software Development Life Cycle)**: 요구사항 분석부터 폐기까지의 생애주기 관리.
- **[[Code Review|Code Review]] & Quality Assurance**: 개인의 실수를 시스템이 잡아주는 프로세스.
- **Technical Debt [[Management|Management]]**: 빠른 배포와 안정성 사이의 트레이드오프를 전략적으로 관리.
- **Automation**: CI/CD, 자동화 테스트, 인프라 관리 코딩(IaC).
## 매 핵심
## ⚠️ 모순 및 업데이트 (Contradictions & Updates)
- "일단 작동하게 만든다"는 해커 정신은 엔터프라이즈 환경에서 "누가 봐도 이해하고 유지보수할 수 있게 만든다"는 엔지니어링 원칙보다 하위에 있다. 최근 AI 보조 도구(Copilot 등)의 도입으로 '작성'의 속도보다는 '검증'과 '설계 역량'이 엔지니어의 핵심 덕목으로 더욱 강조되고 있다.
### 매 challenge
- **Scale**: 매 100s of teams.
- **Compliance**: SOX, HIPAA, GDPR, PCI.
- **Legacy**: 매 monolith + 매 mainframe.
- **Multi-stakeholder**: 매 product + ops + security + legal.
- **Long lifecycle**: 매 10년+.
## 🔗 지식 연결 (Graph)
- Related: Agile-Methodology , Test-Driven-Development
- Concept: [[Technical-Debt|Technical-Debt]]
### 매 modern paradigm
- **Platform engineering**: 매 IDP (Internal Dev Platform).
- **DevSecOps**: 매 security shift-left.
- **GitOps**: 매 declarative infra.
- **SRE**: 매 reliability budget.
- **DORA metrics**: 매 4 key.
- **Team Topologies**: 매 stream-aligned + platform.
## 🤖 LLM 활용 힌트 (How to Use This Knowledge)
### 매 architecture
- **Monolith → Microservice**: 매 strangler.
- **Modular monolith**: 매 alternative.
- **Event-driven**: 매 Kafka, EDA.
- **API-first**: 매 OpenAPI, gRPC.
- **Data mesh**: 매 domain-owned data.
**언제 이 지식을 쓰는가:**
- *(TODO)*
### 매 process
- **Agile / Scrum**: 매 small batch.
- **SAFe**: 매 enterprise scale (controversial).
- **Trunk-based**: 매 modern CI/CD.
- **DORA**: 매 deploy frequency, lead time, MTTR, change fail rate.
**언제 쓰면 안 되는가:**
- *(TODO)*
### 매 응용
1. **Banking**: 매 core system.
2. **Telco**: 매 BSS / OSS.
3. **Healthcare**: 매 EHR.
4. **Government**: 매 procurement.
5. **Insurance**: 매 claim.
## 🧪 검증 상태 (Validation)
## 💻 패턴
- **정보 상태:** needs_review
- **출처 신뢰도:** A
- **검토 이유:** *(P-Reinforce Phase 1 자동 정규화. 본문 검증 필요.)*
## 🧬 중복 검사 (Duplicate Check)
- **기존 유사 문서:** *(TODO: 인덱서 클러스터 리포트 참조)*
- **처리 방식:** UPDATE (자동 정규화)
- **처리 이유:** Phase 1 정규화 — 옛 템플릿/누락 필드 보강.
## 🕓 변경 이력 (Changelog)
| 날짜 | 변경 내용 | 처리 방식 | 신뢰도 |
|------|-----------|-----------|--------|
| 2026-05-08 | P-Reinforce Phase 1 정규화 (frontmatter + 헤더 표준화) | UPDATE | A |
## 💻 코드 패턴 (Code Patterns)
**패턴 1:** *(TODO: 이 프로젝트 컨벤션 반영한 구조 스켈레톤)*
```text
# TODO
### DORA metrics dashboard
```python
def compute_dora(deployments, incidents, period_days=30):
return {
'deploy_frequency': len(deployments) / period_days,
'lead_time_p50_hours': median(d.commit_to_deploy_hours for d in deployments),
'mttr_minutes': median(i.detect_to_resolve_min for i in incidents),
'change_fail_rate': sum(d.caused_incident for d in deployments) / len(deployments),
}
```
## 🤔 의사결정 기준 (Decision Criteria)
### Strangler fig (legacy migration)
```typescript
// 매 facade routes new 의 new, old 의 old
async function getUser(id: string) {
if (await featureFlag('new-user-service', { userId: id })) {
return newUserService.fetch(id);
}
return legacyUserDao.findById(id);
}
```
**선택 A를 써야 할 때:**
- *(TODO)*
### Platform IDP (Backstage)
```yaml
# catalog-info.yaml
apiVersion: backstage.io/v1alpha1
kind: Component
metadata:
name: payments-service
annotations:
backstage.io/source-location: url:https://github.com/acme/payments
spec:
type: service
lifecycle: production
owner: team-payments
system: checkout
```
**선택 B를 써야 할 때:**
- *(TODO)*
### GitOps (ArgoCD)
```yaml
apiVersion: argoproj.io/v1alpha1
kind: Application
metadata:
name: payments-prod
spec:
source:
repoURL: https://github.com/acme/k8s-manifests
path: prod/payments
destination:
server: https://kubernetes.default.svc
namespace: payments
syncPolicy:
automated: { prune: true, selfHeal: true }
```
**기본값:**
> *(TODO)*
### SLO / Error Budget
```yaml
# 매 99.9% SLO → 43.2 min/month error budget
slo:
service: payments
objective: 99.9%
measurement_window: 30d
burn_rate_alerts:
- severity: page
lookback: 1h
threshold: 14.4 # 매 fast burn (consume 1d in 1h)
- severity: ticket
lookback: 6h
threshold: 6
```
## ❌ 안티패턴 (Anti-Patterns)
### Audit log
```typescript
function audit(action: string, actor: string, target: string, metadata: object) {
auditStream.publish({
timestamp: new Date().toISOString(),
action, actor, target,
metadata,
correlationId: getRequestId(),
});
}
// 매 immutable + retention 7y
```
- **[안티패턴]:** *(TODO: 무엇을 하면 안 되는가 + 이유 + 대신 무엇을)*
### Compliance check (PII access)
```python
def access_pii(user_id, requester):
if not has_role(requester, 'pii_reader'):
raise PermissionError()
audit('pii_read', requester, user_id, {})
if requires_purpose(user_id):
return prompt_for_purpose(requester)
return fetch_user(user_id)
```
### Multi-tenancy (Postgres RLS)
```sql
ALTER TABLE orders ENABLE ROW LEVEL SECURITY;
CREATE POLICY tenant_isolation ON orders
USING (tenant_id = current_setting('app.tenant_id')::uuid);
```
### Trunk-based deploy
```yaml
on: { push: { branches: [main] } }
jobs:
deploy:
steps:
- run: npm test
- run: npm run build
- run: deploy.sh staging
- run: smoke-test.sh staging
- run: deploy.sh canary 5
- run: monitor.sh canary 5m
- run: deploy.sh prod
```
### Disaster recovery test
```python
def chaos_dr_test():
# 매 quarterly DR drill
primary_db.simulate_failure()
assert app_reads_from(replica_db)
promote(replica_db)
assert app_writes_to(replica_db)
rollback()
```
### Architecture decision record
```markdown
# ADR-0042: Adopt Kafka over RabbitMQ for event bus
## Context
50 services, growing 5/quarter, current RabbitMQ at 80% capacity.
## Decision
Kafka MSK with mTLS, schema registry, 7-day retention.
## Consequences
+ Replay capability
+ Throughput headroom
- Operational complexity
- Cost +30% Year 1
```
### Team Topologies (boundary)
```yaml
teams:
- name: payments
type: stream-aligned
owns: [payments-service, billing-svc]
- name: platform
type: platform
provides: [k8s, observability, secrets]
serves: [payments, checkout, ...]
- name: security
type: enabling
enables: [...]
```
### AI-augmented dev (Copilot policies)
```yaml
ai_policy:
copilot: enabled
data_residency: eu-west-1
excluded_paths:
- secrets/
- compliance/
audit_log: true
```
## 매 결정 기준
| 상황 | Approach |
|---|---|
| Greenfield | Cloud-native + IDP |
| Legacy modernize | Strangler fig |
| Compliance-heavy | DevSecOps + audit |
| Multi-team | Platform engineering |
| Reliability | SLO + error budget |
| Slow deploys | Trunk-based + CI/CD |
**기본값**: 매 platform IDP + 매 GitOps + 매 SLO + 매 DORA tracking + 매 trunk-based + 매 ADR.
## 🔗 Graph
- 부모: [[Software-Engineering]]
- 변형: [[Platform-Engineering]] · [[DevSecOps]] · [[SRE]]
- 응용: [[Microservices]] · [[Modular-Monolith]] · [[GitOps]]
- Adjacent: [[Team-Topologies]] · [[DORA]] · [[Backstage]] · [[ArgoCD]] · [[Development Communication Standards]]
## 🤖 LLM 활용
**언제**: 매 large org. 매 regulated industry. 매 long-lived system.
**언제 X**: 매 startup MVP. 매 throwaway.
## ❌ 안티패턴
- **Process for process sake**: 매 velocity ↓.
- **Skip compliance**: 매 fines.
- **Big-bang migration**: 매 risk.
- **Single team owns all**: 매 bottleneck.
- **No DORA measurement**: 매 improvement 의 invisible.
## 🧪 검증 / 중복
- Verified (Accelerate, Team Topologies, Google SRE, Platform Engineering).
- 신뢰도 A.
## 🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-04-20 | Auto-reinforced |
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — DORA + 매 strangler / Backstage / GitOps / SLO / RLS code |