VAR Model Predicts Aperiodic Backup Windows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage systems face challenges in scheduling backup operations efficiently, as customers often lack visibility into system resource utilization, leading to potential interference with primary application workloads and storage system background tasks, resulting in suboptimal operation.
Innovation Solution
A multivariate time series model, specifically a Vector Auto Regression (VAR) model, is built to predict aperiodic backup time windows based on fabric, disk, and CPU utilization data, allowing backup operations to be automatically initiated during low-usage periods, thereby minimizing interference with critical workloads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If backup operations are scheduled at fixed intervals, then backup reliability is improved, but system resource utilization deteriorates due to interference with primary workloads and background tasks
Solution Approach 1:
The patent applies dynamics by transitioning from static fixed-interval backup scheduling to dynamic adaptive scheduling. The system continuously monitors system resource utilization metrics (CPU, memory, I/O, storage capacity) and adjusts backup timing based on real-time conditions. This allows backup operations to be scheduled during periods of low resource utilization, avoiding interference with primary workloads and background tasks, thereby resolving the contradiction between backup reliability and system productivity.
Solution Approach 2:
The patent implements feedback mechanisms by continuously monitoring system resource utilization and using this information to adjust backup scheduling decisions. The system evaluates current resource states, predicts future utilization patterns, and makes informed decisions about when to execute backup operations. This closed-loop feedback approach ensures backups are performed reliably while minimizing impact on system productivity.
2Reliability
If backup operations are performed frequently, then data loss prevention is improved, but system performance deteriorates due to resource consumption
Solution Approach 1:
The system dynamically adjusts backup frequency based on real-time resource availability rather than using fixed intervals. When resources are abundant, backups are performed more frequently to enhance data loss prevention. When resources are constrained, backup frequency is reduced to maintain system performance. This dynamic approach resolves the contradiction between data loss prevention and system performance.
Solution Approach 2:
The patent changes the parameter of backup frequency from a fixed value to a variable that adapts based on system conditions. By monitoring resource utilization metrics and adjusting backup frequency accordingly, the system optimizes the balance between data loss prevention (requiring frequent backups) and system performance (requiring less frequent backups to conserve resources).
3Ease of operation
If manual backup scheduling is used, then operational control is improved, but time efficiency deteriorates due to lack of visibility and manual intervention requirements
Solution Approach 1:
The patent implements self-service automation where the backup system autonomously monitors resource utilization, predicts optimal backup timing, and executes backup operations without manual intervention. The system provides visibility into resource utilization patterns and automatically makes scheduling decisions, eliminating the need for manual control while improving time efficiency. This resolves the contradiction between operational control and time efficiency.
Solution Approach 2:
The patent replaces manual mechanical scheduling operations with an automated intelligent system. Instead of manual monitoring and decision-making, the system uses automated resource utilization monitoring, predictive analytics, and intelligent scheduling algorithms to determine and execute backup operations. This substitution eliminates manual intervention requirements while maintaining operational control through automated decision-making.
4Productivity
If backup operations are delayed to low-usage periods, then system performance is improved, but backup timeliness deteriorates
Solution Approach 1:
The system dynamically determines backup timing based on real-time resource utilization patterns rather than using fixed schedules. By continuously monitoring system conditions and predicting low-usage periods, the system identifies optimal windows for backup operations that minimize performance impact while maintaining timely backup execution. This dynamic approach resolves the contradiction between system performance and backup timeliness.
Solution Approach 2:
The patent applies preliminary action by predicting future low-usage periods in advance and proactively scheduling backup operations during these predicted windows. The system analyzes historical and real-time resource utilization patterns to forecast optimal backup timing, allowing backups to be executed timely during predicted low-usage periods before actual usage increases. This prevents both performance degradation and backup delays.
Data Source
AI summary
A multivariate time series model such as a Vector Auto Regression (VAR) model is built using fabric utilization, disk utilization, and CPU utilization time series data. The VAR model leverages interdependencies between multiple time-dependent variables to predict the start and length of an aperiodic backup time window, and to cause backup operations to occur during the aperiodic backup time window to thereby exploit the aperiodic backup time window for use in connection with backup operations. By automatically starting backup operations during predicted aperiodic backup time windows where the CPU, disk, and fabric utilization values are predicted to be low, it is possible to implement backup operations during time windows where the backup operations are less likely to interfere with primary application workloads, or system application workloads that need to be implemented to maintain optimal operation of the storage system.