Read a sample: 1. A Neighborhood Shop Goes Online
On a rainy Saturday, Maya unlocks the front door of Juniper & Pine, the neighborhood shop she owns. Before she has finished arranging a display of notebooks, a customer calls to ask whether a particular blue journal is in stock.
Maya checks a shelf, confirms that she has three, and puts one behind the counter. Later that day, two more customers ask about the same journal. One sends a message through social media; the other walks in after checking the shop’s page and finding no product information.
The shop has loyal customers, but its online presence is little more than an address and a photograph. Maya wants a website where people can browse products and send an inquiry. If the site proves useful, she may add online ordering later.
Her friend Leo, a software developer, offers to help. He suggests running the application on Amazon Web Services, or AWS. Maya has heard that the cloud is a place to put a website, but she has a practical question.
“What does a customer’s phone actually connect to?”
That question is a good place to begin learning cloud architecture. Instead of memorizing service names first, we can follow an ordinary request from a customer to the resources that answer it. Along the way, we will see why the location of those resources matters, what the shop gains by using the cloud, and which responsibilities remain with Maya’s team.
Each architecture story in this book starts with a need rather than a product catalog. A team faces a problem, considers a choice, and accepts tradeoffs. As you read, look for four things:
- The requirement: What does the application need to do?
- The flow: Where does a request or piece of data go?
- The decision: Which kind of resource could meet the need?
- The tradeoff: What does the team gain, and what must it still manage?
At Cloud Practitioner depth, you do not need to build Juniper & Pine’s website to understand its architecture. You do need to recognize the roles of the major components and explain why a team might choose them.
Maya’s first requirement is modest: customers should be able to view a product page from a browser. She does not yet need to design a worldwide ordering system. That allows Leo to start with a simple picture.
A customer named Ana opens a browser on her phone and enters the shop’s website address. Her phone is the client: the device, or more precisely the software on it, that requests a service from another system. A laptop browser can be a client, too, as can an application running on a tablet.
The browser needs to find the system responsible for the website. The Domain Name System, or DNS, helps translate a human-readable domain name into information the browser can use to reach the site. Once it has the necessary address information, the browser sends a request over the network.
For this first version of the story, imagine the application is running on a cloud-based server. The application receives Ana’s request, identifies the product page she asked for, and sends a response. The browser uses that response to display the page.
In simplified form, the path looks like this:
Ana’s browser → DNS lookup → website application in AWS → response to Ana’s browser
The arrows represent communication, not a single physical cable. Networks, routing, and other components are involved between Ana and the application. The diagram leaves those details out because the important first distinction is between the client making the request and the cloud resource responding to it.
If Ana opens the page for the blue journal, the application might also need product information and a photograph. Those pieces of data must be stored somewhere the application can retrieve them. A more complete conceptual flow is:
Browser → application → product data and image → application → browser
The application coordinates the response; storage holds data beyond the moment of a single request. An actual website may send image requests directly to a storage or delivery service rather than passing every image through the application. For now, the simplified path helps us identify the different jobs without assuming that all websites are built alike.
A resource is something the team uses or creates to support the workload, such as a computing resource that runs application code or a storage resource that holds product images. A workload is the collection of components and work that together deliver an outcome—in this case, the shop’s website.
The distinction matters because “put the website in the cloud” does not describe a complete architecture. Leo still has to decide where the code runs, where information is stored, how customers reach it, and who can change it.
Maya asks whether she could simply buy a computer and leave it running in the shop. She could. A business can operate its own physical infrastructure, and that choice may make sense in some circumstances. But it would make Maya’s team responsible for obtaining and maintaining the equipment and arranging the surrounding facilities and connectivity.
AWS offers a different approach. Rather than purchasing all the underlying hardware for the website, the team can obtain cloud resources as needed. AWS operates the underlying infrastructure, while the customer selects and uses services to run a workload.
This changes how the team acquires capacity. If Maya buys a server,…