Download to read offline
Uploaded byJaliya Udagedara
Serverless Computing With Azure Functions
The document discusses serverless computing with Azure Functions. It provides an overview of serverless computing and how infrastructure is abstracted and resources are managed by cloud providers. Azure Functions are introduced as an event-driven and serverless platform for developing code. Triggers and bindings in Azure Functions define how functions are invoked and interact with inputs and outputs. Security, proxies, performance considerations and demos of Azure Functions are also summarized.
More Related Content
![[Struyf] Automate Your Tasks With Azure Functions](https://cdn.slidesharecdn.com/ss_thumbnails/ibzygut2r9mbbgdi0qzh-signature-d4523b3bf020616e67040d2a940ba8bf9aeab187b20da49956548f98de22ea12-poli-180606121949-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
Similar to Serverless Computing With Azure Functions
Serverless Computing With Azure Functions
- 1. Serverless Computing with AzureFunctions Jaliya Udagedara Microsoft MVP (Visual Studio & Development Technologies)
- 2. Line-up • Serverless Computing •Azure Functions • Triggers & Bindings • Authorization • Azure Functions Proxies • Demo • Performance Considerations
- 3. Serverless Computing • RequiresServers • Infrastructure details are abstracted • Cloud provider manages resources • Driven by events and triggers • Pay only for the time your code is running
- 4. Azure Functions BoxOneDrive Twilio DropboxSendgrid Azure Storage Develop your way Local development Event-driven scale Reduced Dev Ops
- 5.
- 6. Triggers and Bindings •Triggers: Defines how a function is invoked • Bindings • Input • Output • return value • out parameter • collector object
- 7. Azure Functions Security •Applies to HTTP Triggers • Supports API Keys • Host Keys • Function Keys •Integrated security with popular OAuth providers
- 8. Azure Functions Proxies •Single API Surface • Request/Response Transformations • Mocking
- 9. Demo Function App usingPortal Function App using Visual Studio Azure Function Proxies
- 10. Performance Considerations • Avoidlong running functions • Cross function communication • Write functions to be stateless • Write defensive functions • Use async code but avoid blocking calls
- 11. References • Demo Code •Official Documentation • Samples and Content • Tools • Azure Storage Explorer • Azure Storage Emulator • Azure Cosmos DB Emulator
- 12.
Editor's Notes
- #2 Pre-check Browser Font Size Visual Studio Font Size Start Cosmos DB local storage browser
- #4 Serverless Computing-misnomer Infrastructure OS and Framework There is zero administrative tasks You just deploy your code (function) and it runs Serverless compute scales quickly (almost instantly) and vastly Automatically scales within seconds Difference between PaaS(Platform as a Service) With PaaS, you still need to select the OS, VM Size, Scaling and pay for that even it’s not being utilized
- #5 Serverless Accelerate development C#/F#/JavaScript Experimental Languages: Python, PHP, TypeScript, Batch (.cmd, .bat), Bash, PowerShell Bring your own dependencies: Nuget and NPM Flexible development Visual Studio VS Code Portal Azure CLI Simplified integration Azure Cosmos DB, Storage, Service Bus, Notification Hubs, SendGrid, Twillio etc.
- #7 Triggers: A function must have exactly one trigger Bindings provide a declarative way to connect to data from within your code. Bindings are optional and a function can have multiple input and output bindings.
- #8 Authorization Types Anonymous Function Admin – Host Keys -> _master OAuth Providers: Azure Active Directory, Facebook, Google, Twitter, and Microsoft Account
- #9 Multiple Azure functions as a Single API Surface
- #11 Avoid long running functions Default 5 minutes, Max 10 minutes Cross function communication Use storage queues for cross function communication Write functions to be stateless Functions should be stateless and idempotent if possible Explain: Idempotent Write defensive functions Query for 10,000 rows in a db Create a queue message for each of those rows to process further down the line What if breaks at 5000 Answer: Track items in a set that you’ve completed. Use async code but avoid blocking calls Avoid getting Result