Cloud Infrastructure
January 31, 2024
Overview
Cloud Infrastructure can be thought of as building blocks, puzzle pieces, or primitives. They connect together in a variety of ways, predominantly based on their categories, such as Compute or Storage.
Cloud Infrastructure can be categorized several ways. The blog post What is the Cloud? examines the differences between the various super-categories of services: Infrastructure-as-a-Service (IAAS), Platform-as-a-Service (PAAS), Software-as-a-Service (SAAS), Function-as-a-Service (FAAS).
There are other ways to break down Cloud Infrastructure. These different categorizations can help to find the best tool for any given task. However, it's important to first examine the full scope of cloud computing, starting from the global scope.
Table of Contents
Categorization: Cloud Infrastructure At Each Level of Scope
The following blog posts already outline Cloud Infrastructure at various Levels of Scope:
Categorization: Traditional
Below are the predominant categories for Cloud Infrastructure in Cloud Platforms.
Compute
Unless you purchase bare-metal machines, what you're purchasing (more like renting) is a part of a machine. In the case of Lambda functions, you're purchasing a part of a machine for a moment in time.
Containers
Storage
Database
These are managed database services that take away the overhead of patching and updates to both the DB engine and its underlying OS.
Migration & Transfer
These are tools for migrating applications, databases, and other workloads to the cloud.
Networking & Content Delivery
Includes services for networking such as virtual networks, load balancing, DNS, content delivery networks (CDN), and network security features.
Developer Tools
Tools for CI/CD (Continuous Integration / Delivery), development, and near-realtime analytics.
Management & Governance
Tools for managing cloud resources, cost monitoring, compliance, and governance.
Media Services
Services for processing, storing, and delivering digital media content, including video streaming and encoding.
Machine Learning
Machine Learning and Artificial Intelligence services, providing tools for model building, training, and deployment.
Analytics
Products for data analytics, data warehousing, big data processing, and business intelligence tools.
Security, Identity, & Compliance
Products for cloud security, identity management, access control, compliance auditing, and encryption.
Application Integration
Products for integrating various applications, data sources, and systems using tools such as message queues.
Internet of Things
Services for connecting, managing, and analyzing data from IoT devices.
Categorization: Consumption Model
On-Demand
Services are billed without long-term commitments using start and stop times.
Reserved Instances
Services are reserved for a specific period with upfront payment for cost savings.
Spot Instances
Services are offered at discounted prices for using excess capacity, using a bidding model.
Serverless
Services are billed without long-term commitments on a per-usage basis.
Next Steps
Next steps might involve learning about these tools in greater depth wherever Cloud Platform training is offered, such as: aws.training. Always remember to start small and simple. Serverless products are always the best for small-scale projects. Once workloads increase, it may make sense to move onto On-Demand, followed by Reserved Instances. Cloud Platforms may have Budget tools to help keep track of costs as they rise alongside complexity. Happy building!