User-impact-weighted SLO 대시보드. 측정 *true * 실제 고객 노출에 의한 99.99 % 가동 시간 - 단지 원료 가용성.

---

원료의 문제

모든 SRE는 공식을 알고 있습니다: (total_time − downtime) / total_time. 그러나 메트릭은 사용자의 3 %에 영향을 미치는 ** 3 AM 정전을 치료 ** 동일하게 ** 피크로드에서 모든 사용자에 영향을 미치는 정전 **.

oncall 플랫폼이 표시되면 99.99%, 그것을 의미했다 customers 경험 99.99% 가용성? 또는 인프라가 시간의 99.99 %를 차지 했습니까?

**slo-impact는 실제 질문에 답합니다. **

---

건축

익지않는 가동 시간은 매 분을 조사합니다. 영향을받은 충격 가동 시간 무게와 피크 트래픽이 있는지 여부.

데모 대시보드

In-repo 데모 수집가 (합성 은행 플랫폼, 라이브 고객 데이터 없음)에서 렌더링. docker-compose up 또는 slo-impact] 으로 시작하고 http://localhost:8080.]

어떻게 작동합니까?

각 사건을 위해 우리는 2개의 여분 차원을 붙잡습니다:

| 차원 | 의미 | |—||| | user_fraction | 사용자 기반의 분수가 실제로 영향을 미치는 (0–1) | | [T6] | 사고시 교통 |

** 사고 당 사용자 충격 가동 중단 시간: **

slo-impact

User-impact-weighted SLO dashboard.
Measure true 99.99% uptime by actual customer exposure — not just raw availability.

---

The problem with raw availability

Every SRE knows the formula: (total_time − downtime) / total_time.
But that metric treats a 3 AM outage affecting 3% of users identically to a noon outage affecting all users at peak load.

When your oncall platform showed 99.99%, did it mean your customers experienced 99.99% availability? Or just that your infrastructure was up 99.99% of the time?

slo-impact answers the real question.

---

Architecture

Raw uptime counts every minute. Impact uptime weights who was affected and whether it was peak traffic.

Demo dashboard

Rendered from the in-repo demo collector (synthetic banking platform, no live customer data). Start it with docker-compose up or slo-impact and open http://localhost:8080.

How it works

For each incident we capture two extra dimensions:

User-impact downtime per incident:

impact_seconds = duration × user_fraction × traffic_weight

User-impact uptime:

user_impact_uptime = 1 − (Σ impact_seconds) / window_seconds

A 3 AM outage at 8% traffic affecting 5% of users contributes only 0.4% of what the same outage would contribute at peak. That gap — between what your infrastructure reports and what customers feel — is the number that matters for SLO integrity.

“If your SLO has never been breached, it isn’t an SLO. It’s a slogan.”

---

Quick start

Option 1 — Docker (recommended, zero setup)

git clone https://github.com/Ajay150313/slo-impact.git
cd slo-impact
docker-compose up

Open http://localhost:8080 — the dashboard runs with built-in synthetic data that mirrors a realistic digital banking platform. No Prometheus required.

Option 2 — pip

pip install slo-impact          # once published to PyPI
# or from source:
pip install -e .

slo-impact                      # starts on http://0.0.0.0:8080

Option 3 — from source (development)

git clone https://github.com/Ajay150313/slo-impact.git
cd slo-impact
python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
uvicorn slo_impact.main:app --reload --port 8080

---

Prometheus integration