1 / 24

From Tiers to Services Without Web Services

From Tiers to Services Without Web Services. Nati Shalom CTO & Founder. Overall Presentation Goal. Understand how to design SOA solutions for applications that require high-performance and low-latency

landis
Télécharger la présentation

From Tiers to Services Without Web Services

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. From Tiers to Services Without Web Services Nati ShalomCTO & Founder

  2. Overall Presentation Goal • Understand how to design SOA solutions for applications that require high-performance and low-latency • Understand how to use tools such as Spring and Space Based Architecture to transform existing application into scalable services in a very simple manner

  3. About GigaSpaces • Founded in 2000 • Provides middleware infrastructure product for applications characterized by • High volume transaction processing and • Very Low latency requirements • Distributed (Grid) Application Server • In Memory Data Grid (Caching) • Compute Grid • Messaging Grid • Customer base • Financial Services • Telecom • Defense and goverment

  4. SOA for High Performance Apps • Need a light approach for services • Small footprint • Embedded within a application • POJO Based • Minimize data transformation overhead • No transformation when possible • Optimized for single VM • No network overhead • Communication by reference passing • High availability/reliability with minimal overhead • Leverage In-memory approach when possible • Built-in support for stateful support

  5. Analytic Application Routing Generate Report Parsing Matching Load Compute Analytics Exiting Tier Based Application: Transactional Applications

  6. From Tier Based to Services Application Centric Service Centric

  7. Service A Service B Service A Service B Messaging/ESB Typical SOA model based on ESB Services can be Java, C++, .Net Inbound Outbound Inbound Outbound Content-Based Routing

  8. Space Based Computing - Intro • Write – writes a data object • Notify – generates an event on data updates Write • Read – reads a copy of a data object • Take – reads a data object and deletes it Read Write + Read == Data Caching! Write + Notify == Messaging! Write + Take == Parallel Processing! Take Notify Write Simple!!!

  9. Space Based SOA – SOA with four verbs Enrichment Service Order Book Service Reconc. Service write write read take notify

  10. SBA – real time SOA for stateful services Processing Unit Processing Unit Shared state to enable stateful services IMDG Parsing Routing Parsing Routing Services can be Java, C++, .Net Processing Grid (Business Logic) Take/Notify Write Take/Notify Write Content-Based Routing Messaging Grid

  11. The Latency Path Processing Unit (VM) Processing Unit (VM) Linear Scalability Service Oriented Order Book Example running on SBA Processing Unit (VM) Enrichment Service Order Book Service Reconc. Service

  12. Benefits • Simple – Simplify the entire architecture not just a single component • Guaranteed linear scalability • Latency – kept low due to reduced network calls and I/O • Throughput – maxed out on per processing unit through parallelization. Total throuput can be easily increased by adding more processing units. • Reliability – single cluster used for all tiers, less moving parts.

  13. SBA – making scalability simple! • The architect view • Scaling become simple – you scale by adding more processing unit • The developers view • Testability – 90% of the functionality can be tested on a single VM • Short learning curve -Development is done in similar fashion to writing to a single server. • External applications view • Can view the entire cluster as if it was a single server when executing queries, registering for events or writing data. • Operation view • Deployment is done through a single command. • Single point of access for monitoring and management

  14. Processed Orders Routing Service Bean Processed Orders Routing Service Bean Order Processor ServiceBean Order Processor ServiceBean NotifyEvent Container NotifyEvent Container Polling Event Container Polling Event Container Write Write Take Take Notify Notify Declarative Spring-SBA - Making Scalability Even Simpler Replication/ Partitioning Replication/ Partitioning Space BUS Space BUS

  15. Declarative SBA Motivation • Simple! • Building a fully fledged distributed application becomes just a matter of wiring simple local beans • Its all declarative through annotation or XML • Non intrusiveness through abstraction • High level abstraction provides higher degree of flexibility to plug-in different solutions from different topologies, technologies and standards without compromising on the least common denominator • Extremely Extensible • Very easy to extend the model and the underlying implementation without changing the application • Reduced learning curve – Its just Spring! • Fit into common development framework and infrastructure • Fit natively to the existing Spring model • Utilizes common practices such as annotation, IoC etc.

  16. Declarative SBA Using Mule-ESB Replication/ Partitioning Order Service Bean Order Service Bean Outbound Inbound Inbound Outbound Replication/ Partitioning Space BUS Space BUS

  17. Benefits of using Mule-ESB • Provides connectivity with other sources such as WS, REST etc. • Support complex workflow scenarios • Performance – Optimized for local VM • POJO Based ESB – Simple, efficient • Fit in natively into Spring environment

  18. SLA Driven Application Service Container • Provide built-in support for deployment of Spring based applications • Virtualize the network and physical resources from the application • Handle FailOver, Scaling and Relocation policies using SLA based definitions. • Provide distributed dependency injection to handle partial failure and deployment dependency. • Provide single point of access for monitoring and management

  19. SLA Driven deployment PU-1 PU-1 PU-1 PU-2 PU-2 PU-(N) PU-(N) SLA Driven Container (GSC) SLA Driven Container (GSC) SLA Driven Container (GSC) GS Manager GSM Deploy/ Undeploy • SLA: • Failover policy • Scaling policy • Ststem requirements • Space cluster topology • PU Services beans definition Management And monitoring Spring configuration

  20. Continues High Availability PU-1 PU-1 PU-1 Fail-Over Failure PU-2 PU-2 PU-(N) PU-(N) SLA Driven Container (GSC) SLA Driven Container (GSC) SLA Driven Container (GSC) GS Manager GSM Deploy/ Undeploy • SLA: • Failover policy • Scaling policy • Ststem requirements • Space cluster topology • PU Services beans definition Management And monitoring Spring configuration

  21. SBA use cases in both analytics and transactional applications

  22. FX Trading • Business Challenge – DB is the FX trading WW leader for the past 5 years, and has no plans to become #2… • Ever increasing trade volumes, and introduction of new currencies (and commodities) • Business Requirement: 10ms end-to-end order processing latency • Comply with orders’ scheduling and auditing regulations • Technical Challenge • Envisioned solution was complex. It was requiring 3 different technologies to build up the new system (Application Server, Messaging and Caching) • Seamless scalability, and HA • Zero impact on existing clients (internal and external) • Persist the workflow to meet regulations w/o affecting throughput • Vendor agnostic APIs for new clients and applications • Results • Classical SBA that results in a linearly scalable application (partitioned by currency-pair) • POC performance: 2.5ms order processing latency vs. 10 ms requirement • Zero changes to clients, writing JMS messages • Heavy use of Spring templates • Central management and provisioning

  23. A Dynamically Scalable Architecture for Data IntensiveTrading Analysis Applications • Most financial organizations today use Excel™ or Reporting Databases as the main trading analysis tools. These are very difficult to scale. • The solution is to create a shared In-Memory Data Grid (IMDG) which stores the trading data in a shared pool of machines. Common data calculation and analysis run on that pool as well, leveraging the available memory and CPU resources. • JavaSpaces is a powerful model for distributed persistence. GigaSpaces is a JavaSpaces vendor providing Enterprise features. • Spring hides the details of the JavaSpaces model, allows effort to be focused on requirements rather than frameworks. Using shared data grid for all users Running analytics close to the data to improve performance and leverage the available resources

  24. Reconciliation calculation

More Related