Layered Application Server Processing via Horizontal OS Fabric
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing application server models fail to simultaneously provide reliability, scalability, high performance, and ease of management for mission-critical applications, especially in extreme transaction processing environments, and are limited by legacy or silo-based architectures.
Innovation Solution
A service-oriented middleware system utilizing data center consolidation, server virtualization, and a horizontal Operating System fabric to dynamically distribute applications across multiple machines and data centers, enabling quick modification and resource allocation based on immediate enterprise needs, and supporting service-oriented architecture (SOA) and mashups.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If traditional application server models (open source or stacked products) are used, then cost and accessibility are improved, but reliability, scalability, and performance for mission-critical applications deteriorate
Solution Approach 1:
The system segments the application server functionality into distinct layers (presentation layer, business logic layer, data access layer) that can be independently deployed, managed, and scaled across multiple physical machines. This segmentation allows the system to maintain reliability through redundancy while preserving cost-effectiveness through modular deployment options.
Solution Approach 2:
The application server is designed as a universal platform that can handle multiple functions including web serving, EJB processing, JSP compilation, and transaction management across distributed nodes. This multi-functionality allows a single system architecture to serve both cost-sensitive deployments and mission-critical requirements by dynamically allocating resources based on workload demands.
2Reliability
If application servers are designed for high performance and reliability, then mission-critical application support is improved, but device complexity and management difficulty worsen
Solution Approach 1:
By dividing the application server into layered components (presentation, business logic, data access), the system reduces complexity at each individual node while maintaining high overall reliability through distributed architecture. Each layer can be independently optimized and managed, preventing complexity from compounding across the entire system.
Solution Approach 2:
The system introduces intermediary components such as connection pools, transaction managers, and deployment descriptors that mediate between applications and the underlying infrastructure. These intermediaries abstract complexity away from individual applications while preserving high performance and reliability through centralized management of shared resources.
3Productivity
If application servers handle increasing numbers of simultaneous users and transactions, then productivity and transaction processing capability are improved, but device complexity and resource management difficulty worsen
Solution Approach 1:
The system segments transaction processing across multiple independent server nodes, each handling a portion of the total workload. This segmentation enables linear scaling of transaction processing capability while keeping individual node complexity manageable. Connection pools and thread pools are also segmented and distributed across nodes to handle increased concurrency without proportionally increasing management complexity.
Solution Approach 2:
The system merges multiple server instances into a unified distributed system that presents a single logical interface to clients. This merging allows the system to handle increased transaction volumes by combining resources across nodes while maintaining simplified resource management through centralized configuration and monitoring of the unified system.
4Stability of the object's composition
If legacy silo-based applications are used, then existing system compatibility is maintained, but adaptability to service-oriented architecture and dynamic modification worsen
Solution Approach 1:
The system segments applications into service-oriented components that can be independently deployed and modified. Each component is encapsulated with well-defined interfaces, allowing legacy applications to maintain their internal composition stability while enabling external adaptability through service exposure. This segmentation facilitates gradual migration from silo-based to SOA architectures.
Solution Approach 2:
The application server is designed as a universal platform that supports both legacy silo-based applications and modern service-oriented architectures simultaneously. It provides universal deployment capabilities that can host traditional monolithic applications while also supporting distributed services, mashups, and dynamic composition, thereby maintaining backward compatibility while enabling forward adaptability.
Data Source
AI summary
A system and method for layered application server processing. Embodiments of the system are particularly useful in extreme transaction processing (XTP) environments to support hundreds of thousands of simultaneous users, and perhaps millions of simultaneous transactions. Applications can be built from service components. Mashups can be performed on applications, services, and data, to create a more dynamic form of application. Data center consolidation and server virtualization techniques can be used to distribute applications over multiple machines or data centers. A horizontal Operating System (OS) fabric is then used to integrate components of the data centers in a dynamic fashion, including for example integrating portal or Service Component Architecture (SCA) components into personality profiles that can be federated over many (e.g. thousands) of Java machines and data centers. The net result is that the system can be quickly and dynamically modified or distributed over different machines and different applications, even over distances of different continents, to respond to the immediate needs of the enterprise.