Resource Recovery Sequence Adjustment for Storage OOS Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In computer resource recovery systems, the mismatch between the allocation and recovery rates of storage resources can lead to out-of-space (OOS) issues, causing inefficiencies and poor user experience due to inaccurate space predictions and time-consuming recovery tasks.
Innovation Solution
A method that determines association relations between used and recovered resources over time, adjusts the recovery sequence of resource recovery tasks to minimize the difference between these relations within a predetermined period, thereby optimizing resource allocation and recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If resource recovery operations are performed as background tasks, then resource recovery can be executed without interrupting system operations, but the recovery process takes a long time and may cause out of space (OOS) issues when allocation rate exceeds recovery rate
Solution Approach 1:
The patent applies dynamics by making the resource recovery process adaptive and adjustable. The system dynamically adjusts recovery parameters such as thread pool size, recovery rates, and priority levels based on real-time system conditions including available space, allocation rates, and service level agreements. This allows the recovery process to accelerate when space is abundant and slow down when space is constrained, resolving the contradiction between continuous operation and recovery time.
Solution Approach 2:
The patent implements feedback mechanisms that continuously monitor system state including available space, allocation rates, and recovery progress. This feedback is used to adjust recovery parameters in real-time, preventing OOS conditions by slowing recovery when allocation exceeds recovery, and accelerating recovery when space becomes available. The feedback loop resolves the contradiction by enabling the system to maintain continuity while adapting recovery speed to prevent time loss.
2Quantity of substance
If resource recovery rate is increased to prevent OOS issues, then available space is maintained better, but system performance may be degraded due to excessive recovery operations
Solution Approach 1:
The patent changes parameters such as recovery thread pool size, recovery rate limits, and priority levels based on system conditions. When available space is low, the system increases recovery parameters; when space is abundant or system load is high, it decreases recovery parameters. This dynamic parameter adjustment maintains available space while preventing performance degradation from excessive recovery operations.
Solution Approach 2:
The patent applies partial action by performing recovery operations selectively rather than maximally. The system recovers resources at a controlled rate that prevents OOS conditions without necessarily maximizing recovery speed. This partial action approach maintains sufficient space while avoiding the performance degradation that would result from excessive recovery operations running simultaneously with system workloads.
3Productivity
If multiple resource recovery tasks are executed concurrently, then recovery efficiency is improved, but the risk of OOS issues increases when allocation rate exceeds recovery rate
Solution Approach 1:
The patent makes the concurrency level of recovery tasks dynamic rather than static. The system adjusts the number of concurrent recovery threads based on real-time conditions including available space, system load, and allocation rates. When space is abundant and load is low, more concurrent tasks are permitted; when space is constrained or load is high, concurrency is reduced. This resolves the contradiction by enabling high efficiency when safe and preventing OOS when risky.
Solution Approach 2:
The patent implements feedback control that monitors space availability and allocation rates during concurrent recovery operations. When the system detects that allocation is exceeding recovery or that space is becoming constrained, it reduces the number of concurrent tasks or pauses recovery operations. This feedback mechanism maintains recovery efficiency when conditions permit while preventing OOS issues when conditions deteriorate.
Data Source
AI summary
Embodiments of the present disclosure provide a method for resource recovery, an electronic device, and a program product. The method may include determining a first association relation between a first quantity of a to-be-used resource in a system and time. The method may further include determining a second association relation between a second quantity of the to-be-recovered resource in the system and time on the basis of a plurality of received resource recovery tasks. In addition, the method may include changing the second association relation by adjusting a recovery sequence of the plurality of resource recovery tasks so as to reduce a difference between the first association relation and the second association relation within a predetermined period. Further, the method may include executing the plurality of resource recovery tasks on the basis of the adjusted recovery sequence.