Topology-Based Failure Impact Determination in Computer Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In managing computer systems, existing technologies fail to adequately distinguish between failed and available devices, leading to unimplementable reallocation plans, undesirable migrations, and unnecessary repairs, particularly when terminal servers fail, impacting the manageability of dependent servers.
Innovation Solution
Expanding the change database to differentiate between 'failed', 'impacted', and 'available' devices, allowing for more informed decision-making in workload migration and resource allocation by marking devices as 'impacted' when their manageability is compromised due to dependencies, thus preventing unwise reallocation and repair attempts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If devices are marked only as failed or available, then the system can detect failures, but it cannot distinguish between failed devices and impacted devices, leading to unimplementable reallocation plans
Solution Approach 1:
The patent segments the failure state into three distinct categories: failed, impacted, and available. This segmentation allows the system to precisely identify not only which devices have failed but also which devices are affected by failures and may have compromised manageability. By dividing the binary failed/available classification into three states, the system can generate more accurate and implementable reallocation plans.
2Productivity
If all non-responsive devices are marked as failed, then the system can simplify failure detection, but it may incorrectly mark impacted devices as failed, leading to unnecessary repairs
Solution Approach 1:
The patent introduces an intermediary classification state called 'impacted' that serves as a buffer between 'available' and 'failed'. When a device does not respond to management requests, it is marked as impacted rather than immediately classified as failed. This intermediary state allows the system to investigate further and determine whether the device is truly failed or merely affected by the failure of another device in its dependency chain, thereby avoiding unnecessary repairs.
3Ease of operation
If the system migrates workload to any available device, then resource allocation can be simplified, but it may migrate to impacted devices with compromised manageability, causing migration failures
Solution Approach 1:
The patent performs preliminary classification of devices into failed, impacted, and available states before executing workload migration. By预先 identifying impacted devices through topology analysis and manageability assessment, the system can exclude these devices from migration targets, ensuring that workloads are only migrated to truly available devices with full manageability, thereby preventing migration failures.
Data Source
AI summary
The present invention provides for indicating devices that are impacted by the failures of another device. Then when allocating workloads to devices, non-impacted devices are given priority over impacted devices as allocation targets for workloads.