Cloud Computing reference data
Cloud service models, providers, and deployment patterns.
Topic guide
What Cloud Computing covers
Cloud-computing reference data covers how computing is rented and run: who manages which layer (IaaS, PaaS, SaaS and their descendants), where it runs (public, private, hybrid and multi-cloud), how workloads are packaged (containers and the tools that orchestrate them) and the serverless model that bills by execution. The service-model names are marketing-adjacent categories rather than strict standards, and vendors blur them, so the useful question is always 'which layer do I manage, and which does the provider manage'.
Key ideas
- Shared responsibility
- The provider secures and operates the layers it manages; the customer secures what it deploys on top. Moving from IaaS to SaaS shifts more responsibility to the provider.
- Elasticity
- Capacity that grows and shrinks with demand and is paid for as used, rather than provisioned for the peak.
- Container
- A packaged process with its filesystem and dependencies, isolated by the host kernel. It is lighter than a virtual machine because it shares the kernel.
- Orchestration
- Automating placement, scaling, networking and recovery of many containers; Kubernetes is the dominant system.
- Serverless
- A model where the provider runs and scales code on demand and bills for execution, so servers still exist but are not managed by the customer.
- Region and availability zone
- Regions are geographic areas; zones are isolated data centers within a region. Spreading across zones protects against a single data-center failure.
- Vendor lock-in
- Dependence on provider-specific services that makes moving workloads costly.
Common pitfalls
- Assuming 'the cloud' removes operational work. Someone still owns security configuration, cost control, backups and reliability design.
- Treating multi-cloud as free resilience. It adds cost and complexity and does not help unless the application is built to use it.
- Equating containers with isolation as strong as virtual machines. A shared kernel is a larger attack surface.
- Ignoring cold starts and per-invocation limits when choosing serverless for latency-sensitive paths.
Using it from an agent
- When asked about a service model, state which layers the customer manages in each, not only the acronym.
- Give provider examples as examples, not recommendations, and note that product names change.
- For pricing questions, say that rates and free tiers change and point to the provider's current pricing page.
Start with: Cloud Service Models · Container Orchestration
Datasets in Cloud Computing
-
Cloud Deployment Models
Models that describe where cloud infrastructure runs.
9 entries -
Cloud Service Models
Service models that define the level of management shared.
9 entries -
Container Orchestration
Tools and concepts for managing containers at scale.
19 entries -
Containerization Concepts
Key concepts and terminology in containerization and container orchestration.
14 entries -
Major Cloud Providers
Leading public cloud computing platforms.
11 entries -
Serverless Architecture Patterns
Common patterns for designing serverless applications.
10 entries -
Serverless Computing
Concepts and terms for serverless architectures.
13 entries
Fetch this category
One request returns the category with all of its dataset summaries. Append
?format=yaml or ?format=toon for the other encodings.
GET https://yjtoon.com/api/category/cloud-computing?format=json