ArcGIS for Server 10.7.1: Is ArcGIS for Server on Amazon Web Services (AWS) PaaS or SaaS?
Not sure if the ArcGIS for Server on Amazon Web Services means that the ArcGIS for Server is already deployed and ready for use and thus this cloud type is considered as SaaS
No installation required—You don't have to install ArcGIS software yourself. Instead, you can use ArcGIS Enterprise Cloud Builder Command Line Interface for Amazon Web Services and sample AWS CloudFormation templates to create an ArcGIS Server site or ArcGIS Enterprise deployment on AWS. Once you create your site, you can immediately connect to it and begin publishing services from a local ArcGIS Desktop client or the ArcGIS Enterprise portal web site.
No hardware infrastructure to maintain—Deploying ArcGIS Enterprise on Amazon Web Services requires no special hardware; you just need to connect to the Internet. After creating your site, you can gain fine-grained management of your server through the AWS Management Console, a web application provided by Amazon. You can log in to your server through Microsoft Windows Remote Desktop Connection or SSH to fine-tune your data, services, and applications.
You are still building/creating your site and managing it.
In this sense, which cloud type is the AWS? Is it PaaS or SaaS?
Which cloud type is AWS? It depends on which product you are talking about, AWS has dozens and dozens of offerings. At a very high level, AWS is primarily an IaaS provider with some PaaS offerings. For example, Amazon EC2 is IaaS while Amazon S3 is PaaS. There are lots of SaaS offerings that exist within AWS environments, but Amazon leaves building and selling those SaaS offerings to its partners/customers. For Esri customers, a notable AWS SaaS is ArcGIS Online.
ArcGIS Enterprise on Amazon Web Services (AWS) is a collection of products that make it easier for you to deploy and manage ArcGIS Enterprise on AWS, the products don't manage your deployments for you. There are numerous companies that provide PaaS and SaaS ArcGIS Enterprise offerings on AWS, including Managed Cloud Services | ArcGIS Cloud Managed Hosting
Thanks Joshua and Ahmad.
As per the picture below, any software installation by the cloud provider means that the cloud is categorized as PaaS or SaaS.
What do you think?
The lines between IaaS, PaaS, and SaaS aren't always hard and fast, and it isn't just about software installation but also software management. For example, when you spin up a Windows Server VM on Amazon EC2, the VM image you spin up was built by Amazon (it is common to use Amazon's images instead of building one's own) but the customer is responsible for managing it, updating it, etc... One could argue Amazon installed the OS, but nearly everyone still considers AWS EC2 as IaaS.
I am not sure if your question is about cloud services in general or ArcGIS Enterprise for AWS. In the context of ArcGIS Enterprise for AWS, it might be Esri's tools that are installing the software but it is still the client/customer installing the software on IaaS, they are just using tools by Esri to simplify the process.
I agree with Josh, my understanding ArcGIS Server on Amazon is Iaas since you will me managing the OS and the software. this is supported by the following blog post.
but your post made me question my understanding and do a little research , it looks like esri considers ArcGIS for Server as Paas!!! I am not sure why or what made them consider it as Paas
The GIS Lounge and GIS Wiki web pages are using a 10-yr old whitepaper, which isn't a good reference for IT.
In terms of ArcGIS Server running in AWS, it would be seen as PaaS to the user who just logs in and publishes services because someone else installed and manages the software, but that someone else is you, not Amazon or Esri. The same system can be viewed as software running on IaaS or PaaS depending on which perspective it is being viewed form.
I totally agree with you, ArcGIS Server running in AWS is Iaas for GIS admins who install , maintain and manage it. and its Paas for publishers , users or even developers who utilize the rest api.