slogo

SLOGo - Go for Service Level Objectives (SLOs)

Go utilities for working with OpenSLO (Service Level Objectives) specifications.

Overview

slogo is a Go library that provides convenient utilities for reading, writing, and validating OpenSLO specification files. It builds on top of the OpenSLO Go SDK to provide a simpler interface for working with SLO definitions in JSON and YAML formats.

Features

• 📖 Read OpenSLO objects from files or readers (auto-detects JSON/YAML format)
• 📝 Write OpenSLO objects to files in JSON or YAML format
• ✅ Validate SLO definitions against the OpenSLO specification
• 🔒 Type-safe Go representations of SLO objects
• 🏷️ Predefined severity and attribute constants
• 📊 Comprehensive SLO ontology for metric labeling and categorization
• 🚀 Production-ready examples for RED, USE, AI Agents, and SaaS metrics

Installation

go get github.com/grokify/slogo

Usage

Reading SLO Files

import "github.com/grokify/slogo"

// Read from file (auto-detects JSON or YAML)
objs, err := slogo.ReadFile("slo-definition.yaml")
if err != nil {
    log.Fatal(err)
}

// Read from io.Reader
objs, err := slogo.Read(reader)
if err != nil {
    log.Fatal(err)
}

Validating SLO Objects

// Validate all objects
if err := objs.Validate(); err != nil {
    log.Fatal(err)
}

Writing SLO Files

// Write as YAML
err := objs.WriteFileYAML("output.yaml")
if err != nil {
    log.Fatal(err)
}

// Write as JSON
err := objs.WriteFileJSON("output.json")
if err != nil {
    log.Fatal(err)
}

Using Predefined Constants

import "github.com/grokify/slogo"

// Severity levels
severity := slogo.SeverityCritical // "critical"
severity := slogo.SeverityHigh     // "high"
severity := slogo.SeverityMedium   // "medium"
severity := slogo.SeverityLow      // "low"
severity := slogo.SeverityInfo     // "info"

// Attributes
attr := slogo.AttrQuery // "query"

Using the Ontology System

import (
    v1 "github.com/OpenSLO/go-sdk/pkg/openslo/v1"
    "github.com/grokify/slogo/ontology"
)

// Create labeled SLO metadata
metadata := v1.Metadata{
    Name:        "api-error-rate",
    DisplayName: "API Error Rate",
    Labels: ontology.NewLabels(map[string]string{
        ontology.LabelFramework:  ontology.FrameworkRED,
        ontology.LabelLayer:      ontology.LayerService,
        ontology.LabelScope:      ontology.ScopeCustomerFacing,
        ontology.LabelAudience:   ontology.AudienceSRE,
        ontology.LabelCategory:   ontology.CategoryQuality,
        ontology.LabelSeverity:   ontology.SeverityCritical,
        ontology.LabelTier:       ontology.TierP0,
    }),
}

Ontology

The ontology package provides a comprehensive labeling system for organizing and categorizing SLOs across multiple dimensions:

• 🔧 Frameworks: RED (Rate/Errors/Duration), USE (Utilization/Saturation/Errors), Custom
• 📦 Layers: Service, Infrastructure, Business, Application
• 🎯 Scopes: Customer-facing, Internal, Business-outcome
• 👥 Audiences: SRE, Engineering, Product, Executive, Customer-success
• 📊 Categories: Availability, Latency, Throughput, Quality, Resource, Engagement, Conversion, Cost
• ⚠️ Severities: Critical, High, Medium, Low
• 📈 Metric Types: Rate, Errors, Duration, Utilization, Saturation, Satisfaction, Stickiness, Retention, etc.
• 🏢 Domains: AI-ML, CRM, SaaS, E-commerce, Fintech
• 🗺️ Journey Stages: Acquisition, Activation, Engagement, Retention, Revenue, Referral

This multi-dimensional taxonomy enables effective filtering, querying, and organization of SLOs across different teams and use cases.

Examples

The examples directory contains production-ready SLO examples organized by monitoring framework and use case:

Infrastructure & Service Monitoring

RED Metrics (4 SLOs)

Request-driven service monitoring for APIs and microservices:

• 📈 Rate SLO - Track request throughput (requests per second)
• ❌ Error Rate SLO - Monitor success/failure ratio with 99.9% reliability target
• ⏱️ Duration SLOs - P95 and P99 latency monitoring for response times

USE Metrics (11 SLOs)

Infrastructure resource monitoring with Brendan Gregg's USE methodology:

• 📊 Utilization - CPU, Memory, Disk space usage
• 🌊 Saturation - CPU load average, swap usage, disk I/O, network bandwidth
• ⚠️ Errors - Disk I/O errors, network errors, memory ECC errors, CPU throttling

Business & Product Metrics

AI Agents (20 SLOs)

Comprehensive monitoring for AI agent platforms with both aggregated and per-user metrics:

• 🟢 Availability - Service uptime and per-user consistency
• ⭐ Quality - User satisfaction, accuracy, hallucination tracking
• ⚡ Performance - Response time, first-token latency
• ✅ Task Success - Completion rates, abandonment, multi-step tasks
• 👥 Engagement - DAU, retention, session duration, conversation depth
• 💰 Cost Efficiency - Token usage, per-user costs, cache hit rates

SaaS CRM (25 SLOs)

End-to-end user journey metrics for CRM platforms (Salesforce, HubSpot):

• 🚀 Activation - User onboarding, time to first value, activation rates
• 💬 Engagement - DAU, MAU, DAU/MAU ratio (stickiness), WAU, power users
• 🔧 Feature Adoption - Contact management, deal pipeline, email integration, reporting, mobile app
• 📈 Business Outcomes - Deal creation, win rates, sales cycle length, contact growth
• 🔄 Retention - Day 7/30 retention, churn rate, cohort analysis, user resurrection

Methodology Examples

Budgeting Methods

Compares different budgeting methods (Occurrences, Timeslices, and RatioTimeslices):

• 🟢 Availability SLO - Internet provider SLA with 99% uptime
• 🔢 Occurrences SLO - Search latency treating all searches equally
• 📊 Ratio Timeslices SLO - Main page availability based on response codes

Treat Low Traffic as Equally Important

Maintains high reliability regardless of traffic volume, ensuring low-usage services receive equal attention.

Example Features

Each example includes:

• ✅ Complete, working Go code
• 📋 Detailed descriptions of what is being measured
• 🏷️ OpenSLO-compliant metadata with ontology labels
• 🔍 Prometheus/BigQuery query examples
• 🧪 Automated validation tests
• 📚 README with methodology explanations and best practices

Project Structure

slogo/
├── ontology/          # SLO labeling and categorization system
├── examples/
│   ├── red-metrics/       # Rate, Errors, Duration (service monitoring)
│   ├── use-metrics/       # Utilization, Saturation, Errors (infrastructure)
│   ├── ai-agents/         # AI agent platform business metrics
│   ├── saas-crm/          # SaaS CRM user journey metrics
│   ├── budgeting-method/  # Budgeting method comparisons
│   └── treat-low-traffic-as-equally-important/
├── datadog/          # Datadog integration utilities
└── cmd/parse/        # CLI tools for parsing SLO files

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Resources

OpenSLO

• 📜 OpenSLO Specification
• 📦 OpenSLO Go SDK

SRE & Monitoring Methodologies

• 📖 Site Reliability Engineering (SRE) Book
• 🔴 The RED Method - Tom Wilkie, Grafana Labs
• ⚡ The USE Method - Brendan Gregg
• 🔔 Google's Four Golden Signals

Product & Business Metrics

• 📈 Measuring Product Health - Reforge
• 💰 SaaS Metrics 2.0 - For Entrepreneurs
• 📊 The Lean Analytics Book - Alistair Croll & Benjamin Yoskovitz