Non-Transparent Bridge Storage Network Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High availability storage systems face challenges in managing condition messages from components, leading to increased workload for storage processors due to the visibility of all devices and potential perturbations from hot-plugging or cabling, which can impact bus enumeration and error handling.
Innovation Solution
Implementing non-transparent bridges strategically within the hierarchy to isolate downstream devices from the upstream path, using a separate transparent switch for daisy-chaining and placing non-transparent bridges downstream of each transparent switch to limit buses visible to the storage processor, while employing a Requester ID Lookup Table to manage device routing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If all storage devices are made visible to the storage processor through transparent bridging, then device accessibility is improved, but system complexity and workload increase due to bus enumeration and error handling
Solution Approach 1:
The patent segments the storage network into multiple domains using non-transparent bridges. Each non-transparent bridge creates a separate domain where devices are not globally visible to all processors, thereby reducing the complexity of bus enumeration and error handling while maintaining device accessibility within each domain.
Solution Approach 2:
The patent introduces non-transparent bridges as intermediary devices between storage processors and storage devices. These bridges act as mediators that control visibility and access, allowing storage processors to access storage devices without direct visibility to all devices in the network, thus reducing system complexity.
2Reliability
If non-transparent bridges are used to isolate downstream devices, then error isolation is improved, but device accessibility deteriorates due to limited visibility
Solution Approach 1:
The patent segments the network into isolated domains using non-transparent bridges, which provide error isolation by preventing errors in one domain from affecting other domains. Simultaneously, transparent switches are used within each domain to maintain full device accessibility for devices that require it, thus balancing error isolation with device accessibility.
Solution Approach 2:
The patent applies different bridging characteristics to different parts of the network: non-transparent bridges are used where error isolation is needed, while transparent switches are used where full device visibility is required. This local differentiation allows the system to optimize for either error isolation or device accessibility depending on the specific network segment.
3Adaptability or versatility
If transparent switches are used for daisy-chaining devices, then device connectivity is improved, but the number of visible buses increases
Solution Approach 1:
The patent uses non-transparent bridges to segment the network at strategic points, which limits the propagation of bus visibility. Transparent switches are used for daisy-chaining within segments, improving device connectivity, but the non-transparent bridges prevent the cumulative effect of increasing visible buses across the entire network.
4Adaptability or versatility
If hot-plugging and cabling operations are performed, then system adaptability is improved, but perturbations occur affecting bus enumeration and error handling
Solution Approach 1:
The patent segments the network using non-transparent bridges, which isolate hot-plugging and cabling operations to specific domains. This segmentation prevents perturbations from propagating across the entire network, thereby maintaining bus enumeration stability in unaffected domains while still allowing system adaptability through hot-plugging operations.
Solution Approach 2:
The non-transparent bridges provide preliminary isolation that prevents perturbations from hot-plugging and cabling operations from affecting bus enumeration and error handling in other parts of the network. This pre-established barrier protects the system's stability against anticipated disturbances.
Data Source
AI summary
A method, computer program product, and computing system for receiving, on a baseboard management controller coupled to an addressable endpoint of a non-transparent bridge included within a storage network, a plurality of condition messages concerning one or more components of the storage network. The plurality of condition messages are filtered into critical condition messages and noncritical condition messages.


