4.9 Consider the impact and requirements of data processing
Hosting, Infrastructure, and Systems
Account for the energy involved in data transfer, considering factors such as the protocol used, whether it is processed client- or server-side, and the environment used.
Criteria
- Batch processing: Human-testable
Use existing and supported carbon-aware computing methods to automate batching and scheduling according to real-time electrical grid carbon intensity data or shift workloads to lower-carbon regions to optimize sustainability while maintaining performance.- Beginner’s Guide to Batch Processing
- GPF – General Policy Framework (PDF) – 6.5 – Front-End (Upload Triggers)
- GPF – General Policy Framework (PDF) – 7.3 – Back-End (Background Processing)
- GPF – General Policy Framework (PDF) – 8.10 – Hosting (Asynchronous Requests)
- GR491 – 3-7043 – Client / Server Requests
- GR491 – 5-6024 – Transfer Compression
- Microsoft Azure WAF – Performance efficiency principles
- Optimizing Data Pipelines
- What is Batch Processing?
- Protocols: Machine-testable
Choose communication protocols appropriate to user needs and the type of data being transferred. Avoid insecure options such as HTTP and FTP, and prioritize secure, efficient alternatives such as HTTPS and SSH. Use modern protocols to take advantage of newer features, while maintaining backward compatibility for older devices.- GPF – General Policy Framework (PDF) – 1.7 – Strategy (Encryption)
- GPF – General Policy Framework (PDF) – 3.3 – Architecture (Protocol Support)
- How HTTPS Works
- HTTP/1 vs HTTP/2 vs HTTP/3
- HTTP/3 From A To Z
- Mediocre Engineer’s guide to HTTPS
- The headers we don’t want
- The HTTP crash course nobody asked for
- Why HTTPS matters
- Why You Shouldn’t Use FTP or HTTP if You Care About Security
- Event-driven architecture: Human-testable
Consider using event-driven architecture and microservices when building products with state changes that do not require full page refreshes. Favor these where they offer a more energy-efficient alternative to traditional APIs based on performance, power, and processing factors. Choose the approach that reduces server workload and environmental impact.- AWS WAF – SUS02-BP06 – Implement buffering or throttling to flatten the demand curve
- AWS WAF – SUS03-BP01 – Optimize software and architecture for asynchronous and scheduled jobs
- Event-driven architecture
- GPF – General Policy Framework (PDF) – 1.7 – Strategy (Encryption)
- GPF – General Policy Framework (PDF) – 3.3 – Architecture (Protocol Support)
- GPF – General Policy Framework (PDF) – 6.5 – Front-End (Upload Triggers)
- GPF – General Policy Framework (PDF) – 7.3 – Back-End (Background Processing)
- GPF – General Policy Framework (PDF) – 8.10 – Hosting (Asynchronous Requests)
- GR491 – 3-7043 – Client / Server Requests
- GR491 – 5-6024 – Transfer Compression
- Microsoft Azure WAF – Performance efficiency principles
- The Complete Guide to Event-Driven Architecture
- What is an Event-Driven Architecture?
- What Is Event-Driven Architecture?
- Client vs server: Human-testable
Avoid redundant processing. When data processing is necessary, carefully compare the relative effects of client- versus server-side processing based on efficiency, performance, security, and sustainability metrics to make an informed decision.- Best practises for 5G App Developers (PDF)
- Client-Side Rendering vs Server-Side Rendering
- Client-Side Vs. Server-Side Testing
- Comparisons of Server-side Rendering and Client-side Rendering for Web Pages (PDF)
- GPF – General Policy Framework (PDF) – 1.7 – Strategy (Encryption)
- GPF – General Policy Framework (PDF) – 6.5 – Front-End (Upload Triggers)
- GPF – General Policy Framework (PDF) – 7.3 – Back-End (Background Processing)
- GR491 – 3-7043 – Client / Server Requests
- GR491 – 5-6024 – Transfer Compression
- Microsoft Azure WAF – Performance efficiency principles
Benefits
- Economic
Improving the efficiency of data processing saves money due to energy and infrastructure needs. - Environment
Running servers for less time reduces carbon emissions. - Performance
Processing data in energy efficient batches can reduce thrashing of hardware during high-intensity periods, maintaining performance stability. - Social Equity
Reducing data processing demand means the resources that data centres demand, and can place a strain on local communities can also be reduced.
GRI
- Materials: Low
- Energy: Low
- Water: Low
- Emissions: Low