Architecting for Scale

What Should Be a Service?
Dividing into Services
Guideline #1: Specific Business Requirements
Guideline #2: Distinct and Separable Team Ownership
Guideline #3: Naturally Separable Data
Guideline #4: Shared Capabilities/Data
Mixed Reasons
Going Too Far
Finding the Right Balance

4. Services and Data
Stateless Services—Services Without Data
Stateful Services—Services with Data
Data Partitioning
Timely Handling of Growing Pains

5. Dealing with Service Failures
Cascading Service Failures
Responding to a Service Failure
Predictable Response
Understandable Response
Reasonable Response
Determining Failures
Appropriate Action
Graceful Degradation
Graceful Backoff
Fail as Early as Possible
Customer-Caused Problems
Summary

Part III: Tenet 3. Organization: Scaling Your Organization for Modern Applications
6. Service Ownership—STOSA
Single Team Owned Service Architecture
Advantages of a STOSA Application and Organization
What Does It Mean to “Own” a Service?
Using Core Teams and Services
Summary

7. Service Tiers
Application Complexity
What Are Service Tiers?
Assigning Service Tier Labels to Services
Example: Online Store
Using Service Tiers
Expectations
Responsiveness
Dependencies
Summary

8. Service-Level Agreements
What Are SLAs?
External Versus Internal SLAs
Why Are Internal SLAs Important?
SLAs for Problem Diagnosis
Performance Measurements for SLAs
Limit SLAs
Top Percentile SLAs
SLA Conditionals
How Many and Which Internal SLAs?
Why Internal SLAs Are Important

Part IV: Tenet 4. Risk: Risk Management for Modern Applications
9. Using Risk Management When Architecting for Scale
Identify Risk
Remove Worst Offenders
Mitigate
Review Regularly
Managing Risk Summary
Likelihood Versus Severity
The Top 10 List: Low Likelihood, Low Severity Risk
The Order Database: Low Likelihood, High Severity Risk
Custom Fonts: High Likelihood, Low Severity Risk
T-Shirt Photos: High Likelihood, High Severity Risk
The Risk Matrix
Scope of the Risk Matrix
Creating the Risk Matrix
Using the Risk Matrix for Planning
Maintaining the Risk Matrix
Risk Mitigation
Recovery Plans
Disaster Recovery Plans
Improving Our Risk Situation

10. Game Days
Staging Versus Production Environments
Staging/Test Environments
Production Environments
Concerns with Running Game Days in Production
Summary

11. Building Systems with Reduced Risk
Technique #1: Introduce Redundancy
Idempotent Interfaces
Redundancy Improvements That Increase Complexity
Technique #2: Understand Independence
Technique #3: Manage Security
Technique #4: Encourage Simplicity
Technique #5: Build in Self-Repair
Technique #6: Standardize on Operational Processes
Summary

Part V: Tenet 5. Cloud: Utilizing the Cloud
12. Getting Started Architecting for Scale with the Cloud
Six Levels of Cloud Maturity
Level 1: Experimenting with the Cloud
Level 2: Securing the Cloud
Level 3: Using Servers and Applications in the Cloud
Level 4: Enabling Value-Added Managed Services
Level 5: Enabling Cloud-Unique Services
Level 6: Cloud All In
Organization Versus Application Maturity Level
Cloud Adoption Mistakes
Trap #1: Not Trusting Cloud Security
Trap #2: Performing Cloud Migration via Lift-and-Shift
Trap #3: The Lure of Serverless—Depending Too Much on the Hype
When and How to Use Multiple Clouds
Defining What We Mean by Multiple Clouds
Which Model? Which Cloud?
The Cloud in Summary

13. Five Industry Trends Changed by the Cloud
What Has Changed in the Cloud?
Change #1: Acceptance of Microservice-Based Architectures
Change #2: Smaller, More Specialized Cloud Services
Change #3: Greater Focus on the Application
Change #4: The Micro Startup
Change #5: Security and Compliance Has Matured
Change Continues

14. Types of SaaS and Tenancy
Comparing Managed Hosting and Different Types of SaaS
Managed Hosting
Multi-Tenant SaaS
Single-Tenant SaaS
Mixing Different Types of SaaS
Common SaaS Characteristics
SaaS Versus Managed Hosting
Summary

15. Distributing Your Application in the AWS Cloud
AWS Architecture
AWS Region
AWS Availability Zone
Data Center
Architecture Overview
Availability Zones Are Not Data Centers
Maintaining Location Diversity for Availability Reasons
AWS—Mapping Availability Zones in Multiple Accounts
Distributing Your Application

16. Managed Infrastructure
Structure of Cloud-Based Services
Raw Resource
Server-Based Managed Resource
Serverless Managed Resource
Implications of Using Managed Versus Non-Managed Resources
Summary

17. Cloud Resource Allocation
Usage-Based Resources Allocation
Allocated-Capacity Resource Allocation
Changing Allocations
Automated Allocation of Resource Capacity
Issues with Automatic Allocation
Dynamic Allocation, Dynamic Cost
Pros and Cons of Usage-Based Versus Allocated-Capacity

18. Serverless and Functions as a Service
Example Application #1: Event Processing
Example Application #2: Mobile Backend
Example Application #3: Internet of Things Data Intake
Advantages and Disadvantages of FaaS
Serverless Hype and the Future of FaaS

19. Edge Computing
Edge Computing Today
Why We Care
What Should Be in the Edge Versus the Cloud?
How Do We Decide? The Driverless Car
Edge Scaling Isn’t the Same as Cloud Scaling
Criteria for Using Edge Versus Cloud
Eight Keys to Success in the Edge
#1: Be Smart About What Goes on the Edge
#2: Don’t Ignore DevOps Principles in the Edge
#3: Nail a Highly Distributed Deployment Strategy
#4: Reduce Versioning as Much as Possible
#5: Reduce Per Node Provisioning and Configuration Options
#6: Scaling Is an Edge Issue, Not Just a Cloud Issue
#7: Nail Monitoring and Analytics
#8: The Edge Is Not Magic
Edge Computing Overall

20. Geographic Impact on Using the Cloud
Cloud Matters Everywhere, But at Different Levels
Replacement Mentality Impacts How You Adopt Cloud
Which Cloud Is Most Important?
Important Technologies Differ
Data Sovereignty Is Universal
My Take

Part VI: Conclusion
21. Putting It All Together
Tenet #1—Availability
Tenet #2—Architecture
Tenet #3—Organization
Tenet #4—Risk
Tenet #5—Cloud
Architecting for Scale

Index