Event-Driven Serverless Architecture - the next big thing in the cloud (Clemens Vasters at CONNECT17)
•
8 j'aime•1,417 vues
Serverless is the newest trend in cloud computing and reflects a change in how applications are hosted. Modularization of solutions has led to the notion of Microservices, and the ‘serverless’ idea allows for hosting such services on top of infrastructure abstractions that eliminate many of the concerns of the common hosting model where applications are explicitly concerned about their hosts and lifetimes. Event-Driven Architecture concepts are ideal to interconnect these ‘serverless’ services and this session will shed light on these trends and usage scenarios.
Recommandé
Recommandé
Contenu connexe
Tendances
Tendances (20)
Similaire à Event-Driven Serverless Architecture - the next big thing in the cloud (Clemens Vasters at CONNECT17)
Similaire à Event-Driven Serverless Architecture - the next big thing in the cloud (Clemens Vasters at CONNECT17) (20)
Plus de Codit
Plus de Codit (20)
Dernier
Dernier (20)
Event-Driven Serverless Architecture - the next big thing in the cloud (Clemens Vasters at CONNECT17)
- 1. #CONNECT17 Event-Driven Serverless Architecture - the next big thing in the cloud Clemens Vasters Lead Architect @clemensv
- 2. “Serverless” (Spoiler: Of course there are servers) (Spoiler: This isn’t new)
- 3. • “Serverless” is PaaS • The promise of “Platform as a Service” is to liberate you from managing low- level infrastructure and focus on apps and data • Virtualization is IaaS • Containers and container orchestrators are a very flexible way to share, configure, and reuse portions of virtualized infrastructure. • Container platforms may form a complete virtualization layer, including software-defined networking. “Serverless” = Platform as a Service (PaaS) “Serverless” means you supply the app Containers may be how “you manage”
- 4. • Managed Cluster • You supply applications that are deployed on a cluster that allows for placement, replication, ownership consensus and management of stateful resources • Managed Middleware • You supply applications that are deployed on sets of independent, “stateless” middleware servers, like web servers or pure compute hosts • Applications may be always-on or start on demand; typically maintain shared cached state and resources • Managed Functions • You supply (small groups of) function implementations that are run for you when triggered by a configured condition. Some PaaS Compute Options
- 6. • … is responsible for holding, processing, and/or distributing particular kinds of information within the scope of a system • … can be built, deployed, and run independently, meeting defined operational objectives • … communicates with consumers and other services, presenting information using conventions and/or contract assurances • … protects itself against unwanted access, and its information against loss • … handles failure conditions such that failures cannot lead to information corruption A “Service” is software that …
- 7. • Defining property of services is that they’re Autonomous • A service owns all of the state it immediately depends on and manages • A service owns its communication contract • A service can be changed, redeployed, and/or completely replaced • A service has a well-known set of communication paths • Services shall have no shared state with others • Don't depend on or assume any common data store • Don't depend on any shared in-memory state • No sideline communications between services • No opaque side-effects • All communication is explicit • Autonomy is about agility and cross-org collaboration Services: Autonomous Entities
- 8. The modern notion of “Service” is not about code artifact counts or sizes or technology choices. It’s about ownership.
- 9. • A system is a federation of services and systems, aiming to provide a composite solution for a well-defined scope. • The solution scope may be motivated by business, technology, policy, law, culture, or other criteria • A system may appear and act as a service towards other parties. • Systems may share services • Consumers may interact with multiple systems System
- 11. • Scales by cloning the app on multiple servers/VMs/Containers • A microservice application separates functionality into separate smaller services. • Scales out by deploying each service independently creating instances of these services across servers/VMs/containers • A monolith app contains domain specific functionality and is normally divided by functional layers such as web, business and data App 1 App 2 App 1 Runtime and Deployment Models Monolith Microservices
- 12. • Single monolithic database • Tiers of specific technologies Monolithic approach Microservices approach • Graph of interconnected microservices • State typically scoped to the microservice • Variety of technologies used stateless services stateless services with separate stores stateful services stateless presentation services State Management
- 13. Thumbnail Service Thumbnail ServicePhoto Share Service Photo Share Service Photo Share Service Photo Share Service Thumbnail Service Photo Share Service Thumbnail SharedLib-v7 Photo Share Service SharedLib-v1 Photo Share Service node.js Thumbnail Service .NET Photo Share Service V1 Thumbnail Service V1 Thumbnail Service SharedLib-v7 Scale Independently Different Technology Stacks Conflicting Dependencies Independent Deployments Thumbnail Service V2 SharedLib-v1 Microservice Architecture Benefits
- 15. Azure Other CloudsOn Premise Data centers Azure Service Fabric Any OS, Any Cloud Dev Box
- 16. Service Fabric Programming Models & CI/CD Azure Other CloudsDev Box On Premise Data centers
- 18. Azure App Services API Apps Easily build and consume APIs in the cloud Web apps Web apps that scale with your business Mobile apps Build mobile apps for any device Logic Apps Automate business process across SaaS and on-premises
- 19. Web Apps Gallery • One-click installation of Wordpress, Joomla, Drupal Auto scaling • Manual • Automatic, based on load or schedule Continuous integration • Integrate w/ VSTS, GetHub, TeamCIty, Hudson, or BitBucket Platform Features Deployment slots • Staged deployment slots • Maximum depends on pricing tier WebJobs • Run program or script on Web App VMs • Continuously or schedules • Scale to handle load Connectivity • Hybrid connectivity • VNet
- 21. Azure Functions platform features
- 22. • A service can be made up of a fleet of independently deployed functions that jointly operate on a shared set of resources • The service interface is made up from the union of the function interfaces • The function interfaces may be a mix of RPC-style call interfaces and event driven ones Microservices and Functions? Function API and Implementation
- 23. Serverless application platform components
- 25. • “REST“ is great for interactively accumulating and acting on state from multiple sources. • Let‘s not pretend all clients are like that – there‘s a lot more • HTTP and RPC are great to obtain immediate answers. • The longer it takes to generate the answer, the more brittle the model becomes Service Communication
- 26. Command Query Job Handover Report Measurement Notification Request Assignment Update Transfer Trace
- 27. Command Query Job Handover Report Measurement Notification Request Assignment Update Transfer Trace
- 28. Intents Facts
- 29. Intents Facts Messaging Expectations Conversations Contracts Control Transfer Value Transfer History Context Order Schema Eventing
- 30. Messaging Eventing A B C A ? ?
- 31. Device Cloud Gateway Device Queues ! ! ! ! Backend Process Iot Hub, Service Bus < > < > Messaging – Example: Device Command
- 32. Backend Process Logic Apps, Service Bus Workflow Manager Step Handler Step Handler Step Handler Step Handler Accounting and Stuff Messaging – Example: Workflow Execution
- 33. Discrete Series Events Independent Report State Change Actionable Time Ordered Context Partitioned Analyzable Report Condition
- 34. • Report independent, actionable state changes to authorized subscribers • “Blob created” • “Sales lead created” • “Order confirmed” • Allows simple, noninvasive, reactive extension of the functionality of a service Discrete Events are an Extensibility Enabler
- 35. Device Collect Readings Ingestor Read Partition Example: Data Series Processing 🕓 Analyzer “Pull” flow towards stateful instance Stateful instances and Actors
- 36. Device Route Alarm ! Distributor Subscrip- tions ! ! ! ! ! L B Example: Discrete Event Handling –Alarms Handlers “Push” flow towards stateless handlers “Serverless” Functions
- 37. Event Grid Sub-second end-to-end latency in the 99th percentile 10,000,000 events per second per region 24-hour retry with exponential back off for events not delivered Push Push * Coming soon
- 38. Platform-level event plane that’s “just there”
- 39. • “Core” functions of services allow for direct interaction: • Commands, Requests, etc. • Extensions react to events emitted by services based on core activities. • Might turn to the emitting service to ask for details or perform actions. Event-Driven: Cores + Extensions Direct Interaction Extension by subscription Messaging Eventing
- 40. • Services: Autonomous software entities grouped around resource ownership and team ownership scopes • Platform-as-a-Service: Hosting options for services with tailored degrees of control for certain scenarios. • Managed clusters: Complex, stateful, high reliability, always-on. • Managed servers: Stateless/shared-nothing, on-demand start. • Managed functions: Event-driven, short-lived • Messaging and Eventing: Communication backplane for services, eventing as extensibility enabler. Summary
- 41. Event Hubs Telemetry stream ingestion Service Bus & Azure Queues Cloud messaging Event Grid Event distribution Logic Apps Workflow and LOB Integration IoT Hub IoT messaging and manage- ment Notification Hubs Mobile push notifications Relay Discovery, Firewall/NAT Traversal Stream Analytics Data Lake Storm Spark Azure Functions Storage Blobs Microsoft Azure
- 42. © 2017 Microsoft Corporation. All rights reserved. The text in this document is available under the Creative Commons Attribution 3.0 License, additional terms may apply. All other content contained in this document (including, without limitation, trademarks, logos, images, etc.) are not included within the Creative Commons license grant. This document does not provide you with any legal rights to any intellectual property in any Microsoft product. You may copy and use this document for your internal, reference purposes. This document is provided "as-is." Information and views expressed in this document, including URL and other Internet Web site references, may change without notice. You bear the risk of using it. Some examples are for illustration only and are fictitious. No real association is intended or inferred. Microsoft makes no warranties, express or implied, with respect to the information provided here.
Notes de l'éditeur
- How to decide Does app require third party components? Do security requirements mandate the use of third party monitoring software? Does app require specific framework configurations? The more customization that is required, the more one moves towards IaaS and away from PaaS.
- Familiar and fast Leverage existing skills in languages, frameworks, etc. Easily add versioning, SSO, monitoring, etc. Gallery contains common building blocks/templates Enterprise grade Securely connect to on-premises resources with AD integration Compliance – ISO, SOC2, PCI Enterprise-level SLAs Global scale Auto scaling within and across different geographical regions
- Bindings abstract away where the data is coming to Logic Apps Programming free paradigm with a visual designer, drag and drop workflows With Connectors to different types of services (eg an FTP server) Talking Points: (Updated on 4/3/17) Lets talk about what really makes up MSFT’s Serverless platform: At the center of the Serverless platform, is our compute offerings: Azure Functions and Azure Logic Apps. Azure Functions is an event based Serverless compute experience that helps you accelerate your development. Run time is open. Function code can be executed in vm or somewhere else, on prem or in clould. Logic Apps is a powerful orchestration tool. It enables building a Serverless app in minutes – by orchestrating multiple functions using a visual workflow tool. Say you have your apps up and running using Serverless. Congratulations! You now need to collect intelligence from different apps across platforms to take actions upon. There are a few essential components which we think are core to building Serverless applications are: Data/ Storage –Functions has triggers and bindings with Azure document DB and Azure Blob storage ** Triggers: Triggers are event responses used to trigger your custom code. They allow you to respond to events across the Azure platform or on premise. ** Bindings: Bindings represent the necessary meta data used to connect your code to the desired trigger or associated input or output data. Messaging such as queues and topics using Azure Service Bus and Azure Event Hubs Integration – that includes core LOB apps and SaaS apps integration via Azure Logic Apps. Intelligence on data and sentiment/ predictive analysis using Cognitive services and Machine learning Conversation as a service – how do we equip developers to build apps that offer an end-to-end experience for their end users – Azure Bot Service offers a Serverless interactive bot experience. More, developers are spending more time writing code that allows them to add huge business impact with Serverless. MSFT offers numerous development tools such as IDE Support for Visual Studio in functions and Logic Apps, enables local development (vs web browser coding environment), visual debugging capability, all with your tools of choice. Lastly, I also want to highlight top scenarios and use cases for Serverless: Real-time Stream analytics: Customers can use Functions to feed real-time streams of data from application tracking into structured data and store it in SQL online. SaaS event processing: Customers can use Functions and Logic Apps to analyze data from an excel file in Onedrive and perform validation, filtration, sorting and convert data into consumable business charts Web app architecture: Used a lot in creating targeted marketing collaterals – when a customer clicks on a webpage, it triggers a webhook, that uses a function to create an ad that matches the customer profile and displays a completed webpage. Real-time bot messaging: When customers send a message to a chatbox, Functions calls Cortana analytics to generate appropriate answers and sends a response back. //from before: Customers have different paths to build a Serverless app – start by building the distributed application components using functions by leveraging the numerous templates and declarative bindings Or Start with the workflow and orchestration of Serverless application using Azure Logic Apps. The visual designer enables developers to quickly and easily author, edit and visualize orchestration of multiple functions and workflow.
- Lets talk about what really makes up MSFT’s Serverless platform: At the center of the Serverless platform, is our compute offerings: Azure Functions and Azure Logic Apps. Azure Functions is an event based Serverless compute experience that helps you accelerate your development. Logic Apps is a powerful workflow and orchestration tool. It enables building a Serverless app in minutes – by orchestrating multiple functions using a visual workflow tool. Say you have your apps up and running using Serverless. Congratulations! You now need to collect intelligence from different apps across platforms to take actions upon. There are a few essential components which we think are core to building Serverless applications are: Data/ Storage –Functions has triggers and bindings with Azure document DB and Azure Blob storage ** Triggers: Triggers are event responses used to trigger your custom code. They allow you to respond to events across the Azure platform or on premise. ** Bindings: Bindings represent the necessary meta data used to connect your code to the desired trigger or associated input or output data. Messaging such as queues and topics using Azure Service Bus and Azure Event Hubs Integration – that includes core LOB apps and SaaS apps integration via Azure Logic Apps. Intelligence on data and sentiment/ predictive analysis using Cognitive services and Machine learning Conversation as a service – how do we equip developers to build apps that offer an end-to-end experience for their end users – Azure Bot Service offers a Serverless interactive bot experience. More, developers are spending more time writing code that allows them to add huge business impact with Serverless. MSFT offers numerous development tools such as IDE Support for Visual Studio in functions and Logic Apps, enables local development (vs web browser coding environment), visual debugging capability, all with your tools of choice. Lastly, I also want to highlight top scenarios and use cases for Serverless: Real-time Stream analytics: Customers can use Functions to feed real-time streams of data from application tracking into structured data and store it in SQL online. SaaS event processing: Customers can use Functions and Logic Apps to analyze data from an excel file in Onedrive and perform validation, filtration, sorting and convert data into consumable business charts Web app architecture: Used a lot in creating targeted marketing collaterals – when a customer clicks on a webpage, it triggers a webhook, that uses a function to create an ad that matches the customer profile and displays a completed webpage. Real-time bot messaging: When customers send a message to a chatbox, Functions calls Cortana analytics to generate appropriate answers and sends a response back. //from before: Customers have different paths to build a Serverless app – start by building the distributed application components using functions by leveraging the numerous templates and declarative bindings Or Start with the workflow and orchestration of Serverless application using Azure Logic Apps. The visual designer enables developers to quickly and easily author, edit and visualize orchestration of multiple functions and workflow.