In what ways is serverless used?
Serverless is a trend in the cloud computing world, but what does it really mean?
Updated
On this page
Updated November 2026. This was first published on 14 February 2022. The idea is the same. Any number below is what the original piece said then. It is not a new count.
Key takeaways
- Serverless is a trend in the cloud computing world, but what does it really mean?
- How does it compare with options offered by Amazon Web Services, Microsoft Azure, and Google Cloud Platform?
- Learn about serverless and functions as a service (FaaS) in this blog post and check out serverless options in Azure, AWS, and Google Cloud.
Serverless is a trend in the cloud computing world, but what does it really mean? How does it compare with options offered by Amazon Web Services, Microsoft Azure, and Google Cloud Platform?
Learn about serverless and functions as a service (FaaS) in this blog post and check out serverless options in Azure, AWS, and Google Cloud.
What does “serverless” refer to? #
Serverless is one of those terms that’s used for all types of things - some of which are real, and others that are marketing fluff. But, at its core, serverless describes a design approach that lets you manage servers asynchronously to run entire applications.
In serverless, there will still be servers. you won’t have to worry about their patching, performance, or fault tolerance.
With serverless apps, we don’t have to worry about scaling up or down, while with traditional server-based applications, we have to ensure there’s always something in place for our users.
Serverless doesn’t require processing unless you’re actively using it, and there’s nothing to pay for if you’re not using it.
As the original public serverless offering, Google App Engine was introduced in 2008, way before Google Cloud Platform existed.
It featured a pay-for-what-you-consume pricing model and let users execute Python code discreetly.
FaaS (Functions as a Service) is a concept #
Serverless architectures are built on functions as a service or FaaS. When an event occurs - such as when a REST API request is received - these services handle the request and perform whatever tasks you program them to do.
GCP calls its FaaS offering Cloud Functions, Azure has Functions, and AWS calls their Lambda. Each of these services supports major languages, including Java, Python, Node.js, and C#. If you cannot find your language, each of them allows you to use your own runtime.
But, the way to accomplish this varies depending on the provider, but you basically create a custom Docker container and configure the Faas platform to use it.
Serverless frameworks - what are they and how do they work? #
The term “serverless framework” is frequently used. Creating a few functions might be straightforward to maintain and deploy. managing a complete application with hundreds of functions can be challenging.
This is where serverless frameworks come in. Serverless frameworks are usually sets of utilities that make managing and deploying large amounts of serverless resources and functions easy.
The Amazon Web Services Serverless Application Model, or SAM, features a cute squirrel mascot.
As it is built specifically for AWS resources, it is very easy to create and deploy functions on AWS. If you’re interested in multi-cloud support, you may want to look into the Serverless Framework.
With support for several clouds, including Azure, AWS, and GCP, it’s quite popular and well-documented.
Others use Architect, Zappa, and Apex Up as well. Pulumi and Terraform are also used for serverless, especially if Infrastructure as Code is already in place.
What is orchestration? #
The presence of a framework and a collection of functions is all well and good. if you want to make serious applications, you need some way of tying the functions together. This is our middleware, allowing us to chain together discrete functions into something that fulfils our needs.
There are some tools that cloud providers have, but you can also use traditional middleware. You can create stateful workflows using Azure Durable Functions. are an extension of Azure Functions, by using orchestration functions and entity functions.
GCP recommends Workflows for serverless orchestration, and Amazon has a similar tool called Step Functions.
Serverless applications, for all their promise, can be difficult to troubleshoot and debug. Debugging serverless components locally before deploying to the cloud is the most popular method. AWS, GCP, and Azure all feature a strong set of tools that allow you to locally simulate just about all components and walk through your serverless code in your IDE.
The challenges increase once you’ve deployed your serverless app to the cloud and are out in the field. You can use AWS X-Ray to analyze traces on your apps in Amazon Web Services to identify performance issues or bottlenecks.
In Azure, Application Insights is the distributed trace tool. It is also possible to look at your serverless app’s logs, both in AWS using Amazon CloudWatch and in Google Cloud Platform using Cloud Logging and Error Reporting. Azure logs can be reviewed in the App Service platform, or they can be sent to Application Insights.
This really is just a quick overview, but hopefully, it offers some insight into how Azure, AWS, and GCP compare when it comes to serverless. For a more in-depth view, check out this comparison of Azure Functions, AWS Lambda, and Google Cloud Functions.
Questions and answers #
What does “serverless” refer to? #
Serverless is one of those terms that’s used for all types of things - some of which are real, and others that are marketing fluff. But, at its core, serverless describes a design approach that lets you manage servers asynchronously to run entire applications.
Serverless frameworks - what are they and how do they work? #
The term “serverless framework” is frequently used. Creating a few functions might be straightforward to maintain and deploy.
What is orchestration? #
The presence of a framework and a collection of functions is all well and good. if you want to make serious applications, you need some way of tying the functions together.