This chapter covers serverless computing as a broader concept, building on the specific Azure Functions chapter with the general principles behind serverless architecture. AZ-900 tests this under objective 1.2.
Jump to a section
A simple way to picture Serverless Computing Concepts
Imagine paying for electricity only for the exact moments a light is switched on, with nothing charged while it's off, and no need to think about the power plant generating that electricity at all. Serverless computing works similarly for code: you pay only while your code is actually executing, and you never have to think about or manage the underlying servers running it.
What serverless computing means
Serverless computing is an approach where the cloud provider fully manages the underlying infrastructure required to run code, automatically allocating resources as needed and scaling down (often to zero) when not in use. The customer focuses purely on their code, not on provisioning or managing servers.
Key characteristics
No server management: the customer never provisions or configures servers directly.
Automatic scaling: resources scale to match demand automatically, including scaling down to zero when idle.
Consumption-based billing: cost is generally tied to actual execution, not to reserved, always-on capacity.
Azure Functions as the primary example
Azure Functions (covered in its own dedicated chapter) is Azure's flagship serverless compute example, but the broader serverless concept can also apply to certain other services that similarly abstract away infrastructure management.
How it fits the broader cloud service model spectrum
Serverless computing represents a further point along the IaaS-to-PaaS-to-SaaS spectrum (covered in the service models chapter) toward abstracting away infrastructure concerns — even more so than typical PaaS, since the customer doesn't think about server capacity at all, even conceptually.
Identify a suitable, event-driven workload
Serverless computing fits workloads that run in response to specific events, rather than continuously active applications.
Write code without provisioning infrastructure
The customer writes and deploys code without needing to provision or configure any underlying servers.
Let the platform scale automatically
The serverless platform automatically allocates resources when needed and scales down when idle, without manual intervention.
A company processing occasional, unpredictable bursts of data uses a serverless approach so they're never paying for idle, always-on infrastructure between those bursts — resources are allocated automatically only when there's actual work to do.
Objective 1.2 expects candidates to understand serverless computing as a broader concept — no server management, automatic scaling, consumption-based billing — with Azure Functions as the primary Azure example.
A common wrong answer is thinking serverless means literally no servers exist — servers are still involved, just not managed by the customer.
Stable terms: serverless computing. Memory trick: serverless = the provider fully handles infrastructure, scaling automatically (including to zero), billed based on actual execution.
Serverless computing means the cloud provider fully manages underlying infrastructure, automatically allocating and scaling resources.
Key characteristics include no server management, automatic scaling (including to zero), and consumption-based billing.
Azure Functions is Azure's primary example of serverless compute.
Mistake
Serverless computing means there are no servers at all.
Correct
Servers are still involved behind the scenes — 'serverless' means the customer doesn't provision or manage them; the provider handles that automatically.
Mistake
Serverless computing is a completely separate concept from PaaS.
Correct
Serverless computing extends the same general idea behind PaaS — abstracting away infrastructure — even further, so the customer doesn't think about server capacity at all.
It means the cloud provider fully manages the underlying servers needed to run code, automatically allocating resources as needed and scaling down when idle — the customer focuses purely on their code, not infrastructure.
It's the primary and most commonly cited example, but the broader serverless concept — automatic scaling, no server management, consumption-based billing — can apply to certain other Azure services that similarly abstract away infrastructure.
You've just covered Serverless Computing Concepts — now see how well it sticks with free AZ-900 practice questions. Full explanations included, no account needed.
Done with this chapter?