Backup Storage Maintenance Management via Dynamic Load Profiling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Maintenance operations in backup storage systems often overlap with critical operations, burdening system resources due to fixed periodic scheduling, which does not account for system idle times.
Innovation Solution
A method that detects triggers for maintenance operations, profiles load factoring features, computes a current load factor, and defers maintenance to later times if the load factor meets a permissible threshold, using a system comprising a data protection agent, load profiling agent, load predicting agent, and maintenance agent.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If maintenance operations are scheduled at fixed periodicities, then maintenance operations can be performed regularly, but they overlap with critical operations and burden system resources
Solution Approach 1:
The patent applies dynamics by transitioning from fixed periodic scheduling to dynamic load-based scheduling. The maintenance operation timing is made adaptive to system conditions, specifically the load factor calculated from multiple profiling features. This allows the system to automatically adjust maintenance timing based on real-time resource availability, resolving the contradiction between regular maintenance and resource availability.
Solution Approach 2:
The patent changes the scheduling parameter from fixed time intervals to variable timing based on load factor thresholds. By introducing a load factor parameter that incorporates multiple system state features, the maintenance scheduling becomes contingent on system conditions rather than predetermined times, enabling maintenance to occur during low-utilization periods without compromising critical operations.
2Productivity
If maintenance operations are deferred to later times, then system resources are preserved for critical operations, but maintenance operations may be delayed excessively
Solution Approach 1:
The patent implements feedback through continuous monitoring of load factoring features and dynamic recalculation of the load factor. This feedback mechanism tracks system utilization over time and provides real-time information about resource availability, enabling the system to identify appropriate maintenance windows while preventing excessive delays through threshold-based triggering.
Solution Approach 2:
The patent applies preliminary action by computing the load factor in advance using multiple profiling features before executing maintenance operations. This proactive assessment of system state allows the system to identify and execute maintenance during predicted low-utilization periods, preventing both resource conflicts and excessive delays.
Data Source
AI summary
Maintenance management on backup storage systems. Specifically, the disclosed method and system derive backup storage system load from a collection of profiled load factoring features. The backup storage system load may subsequently drive whether maintenance operations may be deferred to projected non-peak load times or, alternatively, may be permitted to proceed.