Subnet Routing Failover Using Dual Local Identifiers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication systems lack a mechanism to reconfigure communication routes without interrupting running jobs when cable additions or removals occur, leading to potential disruptions in multi-node communication within and between servers.
Innovation Solution
Implementing a subnet manager that assigns primary and secondary local identifiers to host channel adapters, creating multiple routing paths, and dynamically switches to secondary paths during failure events, ensuring uninterrupted data transfer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single routing path is used for communication, then the system structure is simple, but the system reliability deteriorates when cable failures or additions occur
Solution Approach 1:
The patent divides the routing path into multiple segments by creating primary and secondary routing paths. Each path is associated with specific processing partitions and local identifiers, allowing independent management and failure isolation. When the primary path fails, only the affected segment needs to be switched to the secondary path, rather than reconfiguring the entire routing structure.
Solution Approach 2:
The patent pre-configures both primary and secondary routing paths before failures occur. The secondary path is established in advance with corresponding local identifiers and processing partition mappings, so that when a failure is detected, the system can immediately switch to the pre-prepared secondary path without performing complex real-time routing calculations or job interruptions.
2Adaptability or versatility
If routing paths are reconfigured when cable additions or removals occur, then the system adaptability improves, but job interruption increases
Solution Approach 1:
The patent ensures continuous process routing by maintaining active connections through either primary or secondary paths without interrupting job execution. The switching mechanism operates at the routing layer rather than the application layer, allowing data transfer to continue uninterrupted while the underlying path changes. This eliminates job pausing or termination that would otherwise occur during reconfiguration.
Solution Approach 2:
The patent introduces local identifiers as an intermediary layer between the physical routing infrastructure and the running jobs. These identifiers abstract the underlying path changes, allowing the system to switch between primary and secondary paths while presenting a continuous, uninterrupted interface to the applications. The intermediary layer handles the complexity of path switching transparently.
3Reliability
If multiple routing paths are pre-configured, then the system reliability improves during failures, but the device complexity increases
Solution Approach 1:
The patent applies local quality by associating specific attributes (local identifiers, processing partitions) with each routing path segment rather than managing the entire routing structure uniformly. Each primary and secondary path has its own localized configuration, allowing independent management and switching. This reduces the complexity of managing multiple paths by treating each segment with its specific properties rather than requiring global reconfiguration.
Data Source
AI summary
A system includes a routing fabric and management circuitry. The routing fabric connects processing devices. The management circuitry determines an event associated with a first routing path included in the routing fabric. The management circuitry identifies a processing partition that is affected by the event based on a first local identifier. The first local identifier is associated with the first routing path and is assigned to the processing partition, and the processing partition is associated with one or more of the processing devices. The management circuitry routes one or more processes associated with the processing partition based on a second local identifier. The second local identifier is associated with a second routing path included in the routing path and is assigned to the processing partition. The one or more processes are routed using the second routing path.