Yesterday, we introduced the five essential requirements of edge computing. Today, we'll continue from yesterday's edge computing definition to present the three main service models of edge computing services.
Table of Contents
Service Models
The service models of edge computing are mainly divided into three types
- Infrastructure as a Service (IaaS)
- Platform as a Service (PaaS)
- Software as a Service (SaaS)
We will briefly introduce these in the following sections as IaaS, PaaS, and SaaS. I will reference this in the next section.Translation of the Central University of Finance and Economics network
IaaS
Bare metal infrastructure refers to the integration of physical resources and related management capabilities after physicalization. Through physicalization technologies, computing, storage, and networking resources are abstracted, enabling internal process automation and resource management optimization, and further providing dynamic, flexible bare metal infrastructure services to external users. This layer's consumers utilize processing power, storage space, network components, or intermediate software such as 'bare metal computing resources,' and can also control application systems, storage space, deployed applications, and firewalls, load balancers, etc., but do not control the underlying infrastructure layer directly. Instead, they benefit from IaaS-provided convenience services.
Taking major cloud service providers as examples, the following services all fall under the IaaS category.
- AWS: EC2, VPC
- GCP: GCE
- Azure: VM, Block Storage
PaaS
Platform-as-a-Service (PaaS) provides a development, operation, management, and monitoring environment for cloud applications. It can be described as an optimized 'cloud intermediate software' layer. A well-designed platform layer can meet cloud requirements in terms of scalability, availability, and security. Consumers of this layer can use development tools provided by the platform vendor to build their own applications on top of the cloud infrastructure, potentially having full control over the runtime environment (also including partial control over the underlying infrastructure), but do not control the underlying operating system, hardware, or network foundation.
Kubernetes, commonly seen, is largely categorized under PaaS (Container as a Service), and various managed services such as databases also fall into this category.
Taking major cloud service providers as examples, the following services all fall under the PaaS category.
- AWS: EKS, RDS
- GCP: GKE, Cloud SQL
- Azure: AKS, SQL
SaaS
The 'cloud-native service' is a collective term for cloud-native applications, which are built on foundational infrastructure layers and provided through platform layers, and are delivered to users via network interfaces. These services allow users to access and utilize cloud-native capabilities through multi-cloud networking (endpoints), requiring only a browser or network interface to access—no need to install or upgrade cloud-native software, nor to purchase licenses upfront. Instead, users pay based on actual usage. For cloud-native developers, this enables convenient deployment and upgrades of cloud-native services without managing or controlling the underlying infrastructure layers, such as networks, servers, application systems, and storage.
Most users' cloud-native services are essentially SaaS-based, as the cloud-native services they access through the network interface are fundamentally SaaS in nature.
Below are several examples of SaaS.
- Gmail
- Google Drive
- One Drive
The above are three service models for cloud-native services.
Summary
This time, we introduced the three service models of cloud-native services. Compared to explaining them in abstract terms, providing concrete examples offers much more practical insight.
Tomorrow we will introduce the service models in detail, please stay tuned.
Copyright Notice: All articles in this blog are licensed under CC BY-NC-SA 4.0 unless otherwise stated.