Practical CQRS with Windows Azure CQRS architecture principles are - PowerPoint PPT Presentation

Time-Proven Design Practical CQRS with Windows Azure CQRS architecture principles are based on DDD 2 , Event Lokad Flavor Sourcing and cloud experience. By Cloud Enthusiasts Simplified practical approach to build and Lokad team continuously learns maintain complex cloud solutions with small ff and shares Azure experience teams. via conferences and blogs. With OSS Framework Lokad relies on Lokad.CQRS for complex integrations, high-load data processing and business workflows. 1. CQRS = Command Query Responsibility Segregation http://abdullin.com 2. DDD = Domain Driven Design

Limitations Every approach has these We are still learning Based on the current practical experience Only 3 cloud projects Not all scenarios are covered Limited teaching experience Only with distributed teams in EU/ Russia Startup Survival Mindset Limited resources, low-friction development Unlearning Curve A bit of unlearning is needed to build for cloud CQRS with Windows Azure – http://abdullin.com and http://lokad.com

CQRS Command-Query Responsibility Segregation Design Principle Separation of reads from writes Architecture Model for Cloud Patterns + Design + Project Guidelines Messaging Domain-Driven Design CQRS Event Sourcing CQRS with Windows Azure – http://abdullin.com and http://lokad.com

Simple Organizational Model deriving from Command-Query Load Balance Responsibility Segregation Partition Replicate Decide 1 2 3 Command Query Act Report Native cloud distribution for fault- Event tolerance (cloud is cheap but volatile) and performance. CQRS with Windows Azure – http://abdullin.com and http://lokad.com

Worker Role Windows Azure Azure Fabric Web Role Platform as a Service Platform as a Service Custom VM SQL Azure Pay-as-you-go Cloud MS SQL Elastic Scaling Table Native .NET Azure Storage Queue NoSQL Blob CQRS with Windows Azure – http://abdullin.com and http://lokad.com

Client Web Role, ASP.NET MVC 2/3, WinForms, Send Query WCF REST API, AJAX Commands Views Service Bus Views Azure Worker, Azure Blob/Table Service Publish Storage, NoSQL Events Persistence View Handlers Azure SQL, NoSQL, Azure Worker Role, Azure Blob/Table Service Domain Log Azure SQL, Table Storage CQRS with Windows Azure – http://abdullin.com and http://lokad.com

Lokad.CQRS for Windows Azure Cloud Architecture Guidance CodePlex Project CQRS 1 App Engine for Azure Tutorials Samples Native Azure Queues Article Series Architecture Patterns & Practices Atomic Storage Development and deployment Debugging, upgrades and Streaming Storage maintenance Full auditability Project Testability Scalability Theory ASP.NET MVC Web Client Guidance System.Transactions Decoupled Interface Lokad Feature Studies Lokad.Cloud Support Ask.Lokad Community 1. CQRS = Command Query Responsibility Segregation CQRS with Windows Azure – http://abdullin.com and http://lokad.com

Case 1: Salescast Cloud integration engine capable of processing millions of product references, tailored for large retail networks. Lokad.CQRS significantly simplifies the development while preserving cloud Salescast Web UI Maintenance UI scalability. It allows to have .NET 4.0, Lokad.CQRS Client, .NET 4.0, Lokad.CQRS Client, less than one hour interval Windows Forms ASP.NET MVC 2 between committing code and reliably deploying latest changes into production. Lokad Customer DBs Forecasting API Features: Salescast Server Partner APIs Multi-tenant Lokad.CQRS Worker Tenant-specific ad-hoc Integration Servers Lokad Hub integration logic. Full audit logs. Auto detection of 3 rd party Salescast Store business apps. SQL Azure, Azure Storage API. CQRS with Windows Azure – http://abdullin.com and http://lokad.com

Case 2: Callcalc Email-based forecasting client for call centers. You send an Excel spreadsheet with call volumes and Callcalc replies with forecasts. Lokad.CQRS provides simple and reliable foundation for a CallCalc Web UI Email Interface heavily verticalized solution .NET 4.0, Lokad.CQRS Client, .NET 4.0, Lokad.CQRS Client, IMAP that tailors our raw Forecasting ASP.NET MVC 3 API for the very specific needs of call centers. Lokad Features: Forecasting API IMAP Mailbox Multi-tenant Callcalc Server Multiple calling queues Lokad.CQRS Worker Microsoft Excel Erlang-C staffing Lokad Hub optimization IMAP interface Callcalc Store Azure Blob Storage CQRS with Windows Azure – http://abdullin.com and http://lokad.com

Case 3: Lokad Hub Platform unifying metered pay- as-you-go forecasting subscription offered by Lokad. Lokad.CQRS was key to transition a pre-cloud business app toward a decoupled and efficient design. It provided Hub Customer UI Hub Admin UI additional reliability and .NET 4.0, Lokad.CQRS Client, .NET 4.0, Lokad.CQRS Client, allowed much faster iterations. ASP.NET ASP.NET Features: Multi-tenant. Payment Gateway Full audit logs. Flexible reporting, including Hub Server Geo-Location Lokad Analytics ad-hoc reports and Lokad.CQRS Worker temporal queries. CRM System Reliable failure handling. No-downtime upgrades. Integration with payment, Geo-Location and CRM Hub Data systems. SQL Azure, Azure storage CQRS with Windows Azure – http://abdullin.com and http://lokad.com

Conclusion: Diverse Projects leveraging CQRS + Azure Targeting retailers: Massive Scalability – CQRS Lokad Salescast Complexity – DDD Handling integrations with unreliable legacy Targeting call centers: Parsing Excel spreadsheets on the cloud Lokad Callcalc Blending automation and human support Azure seamlessly integrated into existing IT Centralized SaaS subscription management: Metered pay-as-you-go billing Lokad Hub Kept simple and lean with DDD Full auditability to reduce dev & support friction Same Resources and Experience Same Azure Stack Same Architecture Same Tools Same Teams Same Design Principles Same Scalability CQRS with Windows Azure – http://abdullin.com and http://lokad.com

CQRS Scaling Tackle Complexity Teams Performance DDD 1 UI vs Server Storage + Load Models Domain vs Reports Replication Reliable Integration Familiar IDEs Cloud Elasticity Lower Costs Efficient Competitive Less Friction Development Advantage Faster Iterations 1. DDD = Domain Driven Design CQRS with Windows Azure – http://abdullin.com and http://lokad.com

References on CQRS with Windows Azure CQRS Info: http://cqrsinfo.com/ CQRS Starting Point: http://abdullin.com/cqrs/ Lokad.CQRS for Windows Azure: http://code.google.com/p/lokad-cqrs/ Presentation Rinat Abdullin: http://abdullin.com Lokad SAS: http://lokad.com Questions? DDD/CQRS: http://groups.google.com/group/dddcqrs CQRS + Azure: http://groups.google.com/group/lokad Private: rinat.abdullin@gmail.com CQRS with Windows Azure – http://abdullin.com and http://lokad.com