Further Reading: SLIs, SLOs & Error Budgets

Back to SLIs/SLOs

Site Reliability Engineering (Google SRE Book)

Book: Site Reliability Engineering: How Google Runs Production Systems

Why it matters: The definitive guide to SRE practices, including the original explanation of SLIs, SLOs, and error budgets.

Key Concepts

SLIs (Service Level Indicators): - What to measure - How to measure it - User-facing vs internal metrics

SLOs (Service Level Objectives): - How to set targets - What targets are realistic - How to communicate SLOs

Error Budgets: - How to calculate error budgets - How to use error budgets - Error budget policies

Relevance: Provides the foundational concepts and real-world examples from Google.

Key Excerpts

On SLIs:

"SLIs should measure what users care about, not what's easy to measure. Internal metrics like CPU utilization don't reflect user experience."

Key insight: SLIs must be user-facing. Internal metrics don't tell you if users are happy.

On SLOs:

"SLOs should be achievable but not too easy. They should reflect user needs and business requirements, not just current performance."

Relevance: Explains how to set appropriate SLOs that balance user needs with feasibility.

On Error Budgets:

"Error budgets balance reliability and velocity. When error budget is exhausted, stop feature work and focus on reliability."

Key insight: Error budgets provide a data-driven way to balance feature development and reliability work.

The Site Reliability Workbook

Book: The Site Reliability Workbook: Practical Ways to Implement SRE

Why it matters: Practical guide to implementing SRE practices, including detailed SLI/SLO examples.

Key Concepts

Implementing SLIs: - Step-by-step process - Common pitfalls - Real-world examples

Setting SLOs: - How to choose targets - How to validate targets - How to adjust SLOs

Using Error Budgets: - Error budget policies - How to track error budgets - How to respond to budget exhaustion

Relevance: Provides practical, actionable guidance for implementing SLIs/SLOs.

SLI/SLO Best Practices

Google Cloud Documentation

Documentation: SLI/SLO Best Practices

Why it matters: GCP's recommended practices for defining and using SLIs/SLOs.

Key Practices

1. Start Simple - Begin with availability and latency - Add more SLIs as needed - Don't over-engineer initially

2. Measure What Matters - User-facing metrics - Business-critical operations - Not internal implementation details

3. Set Realistic Targets - Based on baseline performance - Account for improvement over time - Balance user needs and feasibility

4. Use Error Budgets - Calculate error budgets from SLOs - Define error budget policies - Track and monitor error budgets

Relevance: Provides concrete best practices for implementing SLIs/SLOs.

Error Budget Policies

Example Policies

Google's Approach: - > 50% remaining: Normal operations - 25-50% remaining: Reduce risky changes - < 25% remaining: Stop feature work - 0% remaining: Emergency reliability only

Why it works: - Gradual escalation - Clear thresholds - Data-driven decisions

Relevance: Provides a concrete example of error budget policy implementation.

Additional Resources

Papers

"Site Reliability Engineering" (Google SRE Book) - Original SLO concepts - Link

"The Site Reliability Workbook" (Google SRE Workbook) - Practical implementation guide - Link

Books

"Implementing Service Level Objectives" by Alex Hidalgo - Comprehensive SLO guide - Practical examples

Online Resources

Google Cloud Documentation: SLO Documentation - GCP SLO implementation - API and tools

SRE Book: sre.google - Free online version - All chapters available

Key Takeaways

  1. SLIs measure user experience: User-facing metrics, not internal metrics
  2. SLOs balance needs and feasibility: Realistic but not too easy
  3. Error budgets enable velocity: Data-driven balance between features and reliability
  4. Start simple: Begin with availability and latency, expand as needed
  5. Monitor and adjust: SLOs should evolve with system and user needs