Storage Node Placement Using L3 Network Topology
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing distributed, scale-out storage systems fail to utilize network topology information for L3 networks, leading to inefficient and resilient communication between clients and nodes, and are unaware of common failure modes and path efficiencies.
Innovation Solution
A system and method that obtain and store network topology information along with meta-information to determine optimal storage resource placement on nodes, including flash, disk, and network interface types, and periodically scan or trigger updates for efficient resource configuration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If distributed, scale-out storage systems are deployed without utilizing network topology information, then deployment scalability is improved, but communication efficiency and system resilience deteriorate
Solution Approach 1:
The system performs preliminary actions by obtaining and analyzing network topology information before making storage placement decisions. The placement determination module proactively evaluates topology data to identify optimal node locations that account for network paths, potential failure modes, and communication efficiency, rather than making placements reactively without topological awareness.
Solution Approach 2:
The system implements feedback mechanisms by continuously monitoring network topology and using this information to adjust and optimize storage placement decisions. The placement determination module incorporates topology feedback to dynamically evaluate placement options, ensuring that storage nodes are positioned to maximize communication efficiency and system resilience based on actual network conditions.
2Device complexity
If storage placement decisions are made without network topology awareness, then system complexity is reduced, but path efficiency between nodes and clients deteriorates
Solution Approach 1:
The placement determination module serves as an intermediary that bridges network topology information and storage placement decisions. This intermediary component processes topology data, analyzes path efficiencies, and translates topological characteristics into optimized placement recommendations, thereby improving path efficiency without requiring the entire storage system to become inherently more complex.
Solution Approach 2:
The system performs preliminary analysis of network topology to identify optimal placement locations before actual storage deployment. By pre-evaluating network paths, potential bottlenecks, and efficiency metrics using topology information, the system establishes optimized placements that improve productivity without adding operational complexity during normal system operation.
3Ease of manufacture
If conventional node placement methods are used without topology consideration, then ease of installation is maintained, but data distribution and collection efficiency deteriorates
Solution Approach 1:
The system implements self-service capabilities by automatically obtaining, analyzing, and applying network topology information to make storage placement decisions. The placement determination module autonomously evaluates topology data and determines optimal node locations without requiring manual intervention or complex installation procedures, thereby maintaining ease of installation while improving data distribution and collection efficiency through topology-aware automated decision-making.
Data Source
AI summary
The present disclosure relates to device in a network. The device includes an obtaining module, a storage module, and a determination module. The obtaining module is configured to obtain topology information associated with one or more nodes available in a network. The storage module is configured to store the topology information obtained along with meta-information associated with the topology information, if any. The determination module is configured to determine placement of at least storage resources on at least one node selected from the one or more nodes based on the topology information and the meta-information.