Host-Gateway Aggregation for Distributed Computing Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current distributed computing systems face challenges in scalability and management complexity, particularly in cloud-computing facilities, where hierarchical layers of virtualized computing facilities are difficult to manage efficiently, leading to increased overhead and technological constraints.
Innovation Solution
The method involves aggregating host computers into clusters managed by a second-level management server, which are then aggregated into a hierarchical distributed computing system managed by a top-level management server using a host-gateway appliance with host-gateway control logic, providing a subset of native management commands to administrators through a top-level management interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If cloud-computing facilities employ hundreds, thousands, or more networked and geographically distributed multi-processor servers controlled by virtualization technology to produce hierarchical layers of virtualized computing facilities, then computational bandwidth and data-storage capacity are greatly increased, but management overheads and complexities increase to the point that continued upward scaling encounters technological constraints and barriers
Solution Approach 1:
The patent segments the distributed computing system into hierarchical levels: individual host computers, clusters of hosts managed by second-level management servers, and multiple clusters aggregated under a top-level management server. This segmentation allows the system to scale by adding clusters rather than requiring a single management server to handle all hosts directly, thereby reducing management complexity while maintaining increased computational capacity
Solution Approach 2:
The patent introduces second-level management servers as intermediaries between the top-level management server and individual host computers. These intermediary servers manage clusters of hosts locally, reducing the communication overhead and complexity at the top level while enabling scalable aggregation of numerous hosts into the distributed computing system
2Ease of operation
If a top-level management server directly manages a large number of distributed host computers, then system control is centralized, but communications overhead and distributed locking operations increase, degrading performance
Solution Approach 1:
The patent divides the management function into segments: the top-level management server maintains centralized control authority, while second-level management servers handle local cluster management operations. This segmentation reduces communications overhead by localizing management decisions to the cluster level, preventing performance degradation that would occur with direct top-level management of all individual hosts
Solution Approach 2:
The patent adds a hierarchical dimension to the management structure, transitioning from a flat single-level management model to a multi-level hierarchical model. This dimensional change allows centralized control to be maintained at the top level while distributing operational management to intermediate levels, thereby reducing communications overhead and improving overall system productivity
Data Source
AI summary
The current document is directed to methods for aggregating host computers into distributed computing systems and to distributed computing systems created by the methods. In a described implementation, host computers are aggregated into two or more clusters, at a first distributed-computing-system level, each managed by a second-level management server. The two or more clusters are then, in turn, aggregated into a hierarchical distributed computing system managed by a top-level management server. The top-level management server is interconnected to, and accesses, the second-level management servers through a host-gateway appliance that includes host-gateway control logic implemented within a server computer. In order to achieve scalability and efficiency, the top-level management server provides a subset of the native management commands to system administrators and other users who access a management interface provided by the top-level management server.