Distributed Compute Elements in Converged Infrastructure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data center designs face limitations in scalability and flexibility due to the separation of application servers, network nodes, and storage nodes, leading to increased processing loads and coarse-grained scaling, which restricts the ability to efficiently scale out storage nodes and application servers.
Innovation Solution
Implementing a converged infrastructure system with distributed computing elements across application server nodes, network switch nodes, and storage media nodes, where compute elements are distributed to collaboratively control and coordinate data processing functions, enabling flexible scaling and efficient data management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If conventional data center designs with separated application servers, network nodes, and storage nodes are used, then system stability and clear functional separation are improved, but scalability and flexibility deteriorate due to coarse-grained scaling requirements
Solution Approach 1:
The patent segments the traditional separated architecture (application servers, network nodes, storage nodes) into a unified fabric where compute elements are distributed across all node types. This segmentation allows fine-grained scaling by enabling individual nodes to perform multiple functions dynamically, resolving the contradiction between stable separation and flexible adaptability.
Solution Approach 2:
The patent implements multi-functionality by enabling application servers, network nodes, and storage nodes to all participate in data processing functions through distributed compute elements. Each node type can serve multiple purposes within the unified fabric, providing scalability while maintaining system stability through consistent architectural principles.
2Productivity
If additional application servers are added to the computing system, then processing capacity is improved, but processing load on back end storage nodes increases requiring addition of new storage nodes
Solution Approach 1:
The patent merges the functions of application servers and storage nodes into a unified fabric where both node types participate in data processing. When application servers are added, the distributed compute elements across all nodes (including existing storage nodes) handle the increased load, eliminating the need to proportionally increase storage nodes and resolving the contradiction between processing capacity and storage node quantity.
3Ease of manufacture
If conventional CI frameworks are used to consolidate systems, then TCO reduction is improved, but scaling flexibility deteriorates due to centralized IT management limitations
Solution Approach 1:
The patent implements dynamics by allowing the system to adapt its architecture dynamically - nodes can assume different roles and functions based on workload requirements. This dynamic capability provides scaling flexibility while maintaining the consolidation benefits of CI frameworks, as the system can reconfigure itself to meet changing demands without requiring complete system redesign.
Data Source
AI summary
Systems, methods, and articles of manufacture comprising processor-readable storage media are provided for implementing converged infrastructure systems with computing elements distributed over server nodes, network nodes, and storage media nodes. For example, a computing system includes application server nodes, network switch nodes, and storage media nodes. The storage media nodes include persistent storage elements to store data. The network switch nodes are configured to enable communication between the application server nodes and the storage media nodes, and to enable communication between the storage media nodes. The application server, network switch, and storage media nodes are configured to implement a converged infrastructure, wherein the converged infrastructure includes compute elements distributed over the application server, network switch, and storage media nodes. The distributed compute elements are configured to collaboratively control and coordinate data processing functions within the converged infrastructure.