Service-oriented architecture

A modular style to software development is based on the use of distributed, loosely coupled, replaceable components.

Transcript

1. Service-oriented architecture @H2Russia internal workshop
2. http://geek-and-poke.com
3. Monolithic Rails applications End user Domain logic Database
4. Issues monolithic applications • Complexity • Hard deploy • Hard use new tools (DB, languages, etc.) • Scale team size • Changes are typically expensive
5. What is it SOA? • A modular style to software development is based on the use of distributed, loosely coupled, replaceable components. • The fundamental concept of an SOA is that each component of the system is broken up into network-accessible services, which are integrated to make a functioning application.
6. Rails application with multiple services End user High level domain logic Service 1 Domain logic Database Service 2 Domain logic Database Service N Domain logic Database
7. Service-Oriented Design with Ruby
8. SOA advantages over monolithic application • Isolation • Robustness • Scalability • Agility • Interoperability • Reuse
9. Isolation • Object-oriented design • Business logic • System Isolation
10. Robustness • Implementation can be changed without the API consumer’s knowledge. • Example: Changing libraries, database or even languages.
11. Scalability • Easy to scale portions of application individually. • Can be handled on a case-by-case instead of requiring optimization of a single database for all cases. • Easier to scale team size.
12. Agility • New services can be implemented outside the full architecture. • Service interface should be versioned. “/api/v1/users/:id” • Design includes the ability to run multiple versions of a service simultaneously. “/api/v1/users/:id” and “/api/v2/users/:id”
13. Interoperability • Prevents being tied to a specific implementation. • Services ease interoperation with internal and external systems and with systems written in languages other than Ruby.
14. Reuse • Service-oriented design enables reuse of components across multiple applications or clients. • If services are created for internal use, they can be exposed later to the public.
15. Converting to Services • Partitioning on Iteration Speed • Partitioning on Logical Function • Partitioning on Read/Write Frequencies • Partitioning on Join Frequency
16. What we need to learn • RESTful-Oriented Architecture • HTTP Methods • HTTP Caching • MIME types • I/O, Threading, and Parallelism • Amazon AWS • Create clients • Testing and Mocking Service Calls • Load Balancing • Security • Web Hooks • Chef • Typhoeus
18. Load Balancing • Weighted Balance • Round-Robin • Least-Connections • URI-Based • Priority • Lowest Latency
19. Caching • HTTP cache (cache-control, ETag, last- modified, expires) • Memcached • Squid and Varnish
20. Debug • Efficiency 2.0 Ops Middleware - track distributed transactions across its Ruby- based service oriented architecture • Monitoring tool.
21. Development • Use stub mode in clients in development environment. • Disable stub mode when develop service.
22. Tests • Test client with service • Test application with stubbed service • Test the full stack
24. @usubov shaliko@ezid.ru Shalva Usubov