4Developers 2019:Serverless approach in Azure - another buzzword or 'Holy Grail'(...)? Paweł Fiderek

### Overview In this talk from 4Developers 2019, Paweł Fiderek explores whether the serverless architecture pattern is merely an overhyped buzzword or a true "Holy Grail" for developers. Using Gartner's hype cycle as an analytical framework, the speaker reviews the historical evolution of hosting and cloud infrastructure, explains key concepts behind Backend as a Service (BaaS) and Function as a Service (FaaS), and demonstrates how to build an event-driven serverless workflow in Microsoft Azure.

### Key Topics - **Hype Cycles and Technology Adoption**: The speaker applies the hype cycle model (technology trigger, peak of inflated expectations, trough of disillusionment, slope of enlightenment, and plateau of productivity) to explain how serverless moved past initial inflated hype into a pragmatic, selective architecture pattern. - **Evolution of Hosting Models**: The transition of modern application deployment from on-premises hardware and shared hosting to virtualization, Infrastructure as a Service (IaaS), Platform as a Service (PaaS), Software as a Service (SaaS), BaaS, and FaaS. - **Shift from Infrastructure to Innovation**: Referencing Mark Russinovich, the presentation highlights how moving higher up the abstraction stack allows organizations and developers to focus on business logic and customer experience rather than server patching, runtime configuration, and operating systems. - **Core Azure Serverless Components**: An introduction to Azure's key building blocks for event-driven systems, including Azure Functions, Azure Logic Apps, Azure Event Grid, Azure Service Bus, and Azure Cosmos DB.

### Technical Details The speaker conducts a hands-on live demonstration implementing an automated employee onboarding notification system using Azure services: - **Azure Functions**: An HTTP-triggered Azure Function configured with anonymous authentication. When invoked with an HTTP POST containing employee details (first name, last name, department), the function generates an email address and unique identifier. - **Azure Cosmos DB**: Used as the persistence layer. A globally distributed multi-model document database where the Azure Function writes the newly generated employee document using native input/output bindings. The speaker notes its multi-model capabilities (key-value, column, graph via Gremlin) and automatic replication across data centers. - **Azure Service Bus**: Configured with a simple queue (`employee-info`). The Azure Function outputs a message onto this queue to signal that a new employee record was created. - **Azure Logic Apps**: Orchestrates the downstream workflow by listening to the Service Bus queue. Because queue messages arrive base64-encoded, Logic Apps decodes the payload, parses the JSON structure without a pre-defined schema, checks conditional logic (whether social sharing is flagged as true), and utilizes built-in social connectors to publish announcements directly to LinkedIn and Twitter.

### Takeaways Serverless is neither an empty buzzword nor an all-encompassing "Holy Grail." It represents a middle ground that provides immense developer productivity and operational offloading for specific workloads, provided teams evaluate pricing models (such as Consumption vs. App Service plans) and operational constraints.