Skip to content

NthLayer

Reliability at build time, not incident time.

Shift reliability left into your CI/CD pipeline. Validate before deploy, not after incidents.

The Reliability Pipeline

   service.yaml
        │
        ▼
┌───────────┐     ┌───────────┐     ┌───────────┐
│  Generate │ ──▶ │  Validate │ ──▶ │  Protect  │ ──▶ Deploy
└───────────┘     └───────────┘     └───────────┘
   apply            verify           check-deploy
   init             validate-spec    portfolio
                    --lint
Stage What Happens Exit Code
Generate Create dashboards, alerts, SLOs from YAML -
Validate Verify specs, lint PromQL, check contracts 1 if invalid
Protect Block deploys when error budget exhausted 2 if blocked

The Problem

Teams deploy code without reliability validation:

  • Alerts created after the first incident
  • Dashboards built after users complain
  • SLOs defined after budget is exhausted
  • No gates to prevent risky deploys

The Solution

NthLayer shifts reliability left - from incident response to CI/CD:

service.yaml
name: payment-api
tier: critical
type: api
dependencies:
  - postgresql
  - redis

# Generate → Validate → Protect → Deploy
nthlayer apply service.yaml --lint
nthlayer verify service.yaml --prometheus-url $PROM_URL
nthlayer check-deploy service.yaml --prometheus-url $PROM_URL
kubectl apply -f generated/

What Gets Generated

Output Description
Dashboards Grafana dashboards with technology-aware panels
Alerts Prometheus alert rules with best-practice thresholds
SLOs OpenSLO-compatible definitions with error budgets
Recording Rules Pre-aggregated metrics for performance
PagerDuty Teams, schedules, and escalation policies

Key Features

Deployment Gates

Block deploys when error budget is exhausted:

check-deploy demo

nthlayer check-deploy service.yaml --prometheus-url $PROM_URL
# Exit code: 0=approved, 1=warning, 2=blocked

SLO Portfolio

Track reliability across your entire organization:

portfolio demo

nthlayer portfolio --path services/

Contract Verification

Verify declared metrics exist before deploy:

verify demo

nthlayer verify service.yaml --prometheus-url $PROM_URL

23 Technology Templates

Pre-built monitoring for:

  • Databases: PostgreSQL, MySQL, MongoDB, Redis, Elasticsearch
  • Message Queues: Kafka, RabbitMQ, NATS, Pulsar
  • Proxies: Nginx, HAProxy, Traefik
  • Infrastructure: Kubernetes, etcd, Consul

CI/CD Integration

.github/workflows/deploy.yml
jobs:
  deploy:
    steps:
      - name: Validate Specs
        run: nthlayer validate-spec services/

      - name: Generate & Lint
        run: nthlayer apply services/*.yaml --lint

      - name: Verify Metrics
        run: nthlayer verify services/*.yaml --prometheus-url $PROM_URL

      - name: Check Deploy Gate
        run: nthlayer check-deploy services/*.yaml --prometheus-url $PROM_URL

      - name: Deploy
        if: success()
        run: kubectl apply -f generated/

Integrates with GitHub Actions, GitLab CI, ArgoCD, Tekton, and Jenkins.

Get Started

  • Installation

    Install NthLayer with pip in 30 seconds

    Install

  • Quick Start

    Generate your first dashboard in 5 minutes

    Quick Start

  • Validate

    Catch issues before they reach production

    Validation

  • Protect

    Block risky deploys with error budget gates

    Protection

The Google SRE Connection

NthLayer automates concepts from the Google SRE Book:

SRE Concept Manual Process NthLayer Automation
Production Readiness Review Multi-week checklist nthlayer verify in CI
Error Budget Policy Spreadsheet tracking nthlayer check-deploy gates
Release Engineering Manual runbooks Generated artifacts + GitOps
Monitoring Standards Wiki pages service.yaml spec