What is Serverless?

Serverless is a computational model widely used in the creation of modern web applications. Understand how the technology works and the advantages of adopting it.

Paulo Moura - Technical Researcher
What is Serverless?

Serverless (or serverless computing) is a software development and execution model where the cloud provider is responsible for executing a piece of code (or function), allocating resources dynamically, which allows developers to build code without worrying about operational tasks like infrastructure provisioning and network configuration. The term “serverless” means that the servers are “invisible” to the developer.

Traditionally, web applications are built and implemented on a server managed by the company itself. Although there is more control over infrastructure resources with this type of approach, this entails a series of responsibilities that impact costs, performance, and productivity, such as:

  • keep the server on the air—even when idle;
  • maintain the server and all its resources;
  • apply appropriate security updates;
  • provision hardware resources to meet demands.

One way to get these challenges out of the way is to migrate to serverless. With serverless computing, everything related to the infrastructure is the responsibility of the service provider, while organizations replace resource provisioning with a payment model in which consumption is based on the number of computational resources used to deliver each user request.

How Serverless Computing Works?

The serverless architecture comprises a set of services that encompasses, among other components: Function-as-a-Service, database and object storage, event transmission and messaging, and API gateway.

Below, we explore in more detail what each of these components does.

Function-as-a-Service (FaaS)

FaaS is a cloud service that executes code in response to events or requests whereby the developer implements code without interacting with physical hardware. Through FaaS, costs are calculated only when resources are actually used, and scalability occurs automatically.

Database and Object Storage

A serverless database combines the functionalities of a traditional database and the characteristics of the serverless architecture, such as automatic and on-demand scalability, consumption-based payment, and infrastructure managed by the provider. On the other hand, the role of object storage is to handle large amounts of unstructured data—something that traditional approaches are unable to do.

Event Transmission and Messaging

Transmission and messaging of data among their components works through distributed data transmission platforms, such as Apache Kafka. This is where the events are received and transmitted both to the database, where they are stored, and to the FaaS service, through which the trigger is pulled so that each function is invoked within the parameters of the respective request; if any failure occurs at this stage, the event is saved in a storage queue to be executed later.

API Gateways

An API gateway acts as a reverse proxy to accept all requests for APIs. User authentication, routing, and rate limiting are examples of everyday tasks performed by API gateways, among other essential management functions that enable infrastructure provisioning in a serverless architecture and resources for integration and interconnectivity via API.

Pros and Cons of Serverless

Now that we know the main components of the serverless architecture, how about we explore the pros and cons that must be considered when deciding on migration? First, let’s see what the main factors are.


Increased developer productivity

Increased productivity is one of the key benefits of serverless. Let’s imagine your team of devs as a touring band. If the members need to transport and install the equipment before each show, this obviously affects the entire preparation. But if the band has the support of a backstage team, they can focus only on the performance, which improves the final result. The same happens with developers, who start concentrating all their effort and talent on development without being distracted by hardware issues.

Developer Experience

Migrating to serverless is also a big step in optimizing the developer experience and thereby attracting and retaining talent for the company. Studies reveal that 53%[1] of professionals in the field prioritize developer experience when analyzing a job opportunity.

In addition to saving developers from interactions with the infrastructure, serverless computing allows the use of any programming language or framework. It also enables quickly carrying out deployments and software updates, since it’s not necessary to promote changes throughout the application but rather to update the function individually and in isolation.

Payment for Execution Only

Unlike other compute models, where the customer pays for the computational resources needed to execute their applications, in the serverless model, the counting starts when a request is made and ends when the execution is completed. Therefore, when no request reaches the server, the client pays absolutely nothing.


In serverless computing, instead of monolithic applications, developers start to build microservices, which are composed of functions that perform a single task individually and independently. Some of the benefits of microservices are:

  • fault tolerance;
  • shorter development cycles;
  • ease of deployment;
  • reuse of codes.

Learn more about microservices and the difference compared to monolithic architectures in this blog post.


Vendor Lock-In

Technological entrapment is a counterpoint to one of the greatest advantages of serverless. Developers focus on what they do best and, on the other hand, some providers don’t offer a platform with open standards, so that the team has to rewrite code—totally or partially— to run the same service with another provider.

Monitoring and Debugging

Microservices architecture contributes to build highly complex applications. If the team doesn’t have solutions that facilitate the monitoring or debugging of existing flaws or vulnerabilities, the complexity itself becomes a disadvantage.


One of the main drawbacks of serverless is latency, resulting from a process known as cold start—learn more about the relation between cold start and serverless in this article. This happens when a function is initialized from scratch to fulfill a request, resulting in a drop in performance and delay in execution, which may compromise the user experience.

Fortunately, serverless computing has advanced considerably in recent years to the point where those limitations mentioned above, which exist largely because of the limitations of cloud computing, are no longer an obstacle. Next, we’ll talk about serverless in its evolved form: edge serverless.

What is Edge Serverless?

When we talk about the digital future and all the requirements it reserves for web applications, the time comes to think about how to move serverless computing to the next level since cloud computing has limitations that have become more evident as the applications modernized. To take this step forward, organizations are complementing their infrastructures with edge computing.

Basically, edge computing takes computing to the closest possible location to the end device—which can be a multitude of things, such as smartphones and IoT sensors. Therefore, edge serverless consists of the execution of serverless functions in a highly distributed infrastructure, which is essential for mission-critical applications, which must deliver better performance and maximum availability, and also applications that require minimum levels of latency.

The Azion Edge Platform allows the deployment of serverless codes in a wide edge network. With WebAssembly support, coding can be optimized by a set of edge-native solutions capable of bringing application development to state of the art in terms of reliability, performance, security, availability, and observability. You can test the platform for free here.

In addition to the possibility of building high-performance applications, the use of edge serverless results in an expanded IT infrastructure on a global scale and NoOps at a significantly lower cost compared to any other computational model. In fact, the increase in revenue achieved by delivering faster solutions and services tends to easily exceed the investment in edge serverless since 59%[2]of customers are willing to pay more for a great experience.

How Azion Helps Developers Get the Most out of Serverless Computing?

The Azion Edge Computing Platform allows organizations to take advantage only of the benefits of serverless computing with a great increase in performance. One of the reasons for this is our globally distributed edge network, which covers even regions where network infrastructure is hostile and difficult to reach, and guarantees the delivery of requests with high performance anywhere.

More than enabling the implementation of serverless applications that require minimum latency levels, Azion uses open standards that enable integration with any cloud provider, eliminating the vendor lock-in problem that exists in many serverless vendors. Therefore, if your company makes use of multi-cloud, your team will be able to develop applications with even greater agility.

Finally, our platform offers a set of observability solutions that facilitate application monitoring by providing detailed event data, which can be transmitted directly to your analysis platform in real time. See how GetNinjas, the largest service contracting site in Brazil, reached unique discoveries with Azion.

Subscribe to our Newsletter