Network Informed Policy Creator for Backup Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current backup software treats all stages of data backup as a single process, leading to unpredictable load distribution and potential failures in primary and secondary storage systems, as they do not effectively determine the best time and path for data transfer based on network conditions and historical data.
Innovation Solution
A network informed policy creator system that decouples backup stages, monitors network device information, and recommends optimal transfer times and paths by analyzing historical data and available bandwidth to ensure efficient data transfers across the network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If backup software treats all stages as a single process, then the backup operation is simplified to one unified process, but the system becomes unpredictable and prone to overloading primary and secondary storage systems
Solution Approach 1:
The patent segments the backup process into distinct stages: data collection stage, data transfer stage, and data restoration stage. Each stage is independently managed with its own policies and parameters, allowing for granular control and optimization of each phase without affecting the entire backup operation.
2Productivity
If backup operations are performed without considering network conditions, then backup processes can be executed continuously, but network congestion occurs and service level objectives are not met
Solution Approach 1:
The patent implements feedback mechanisms that continuously monitor network conditions, system load, and performance metrics. Based on this feedback, the backup system dynamically adjusts transfer rates, schedules operations during optimal times, and modifies resource allocation to prevent network congestion while maintaining high backup productivity.
Solution Approach 2:
The backup system transitions from static, predetermined schedules to dynamic, adaptive operations. Transfer rates, timing, and resource allocation are continuously adjusted based on real-time network conditions and system state, allowing the system to optimize performance and avoid congestion adaptively.
3Productivity
If multiple backups are performed simultaneously from multiple sources, then backup throughput is increased, but network bottlenecks occur on shared routes
Solution Approach 1:
The patent applies different policies and parameters to different backup operations based on their specific characteristics, such as data size, priority, and network path. Each backup operation receives optimized resource allocation tailored to its requirements, allowing high throughput for critical backups while managing bandwidth consumption for less urgent transfers.
Solution Approach 2:
The system allows multiple backup operations to proceed simultaneously but controls their resource consumption through rate limiting and priority-based scheduling. Critical backups receive full bandwidth allocation while non-critical backups are throttled or scheduled for off-peak times, maintaining overall throughput while preventing network saturation.
4Ease of operation
If backup policies are applied uniformly across all stages, then policy management is simplified, but optimal transfer times and paths cannot be determined
Solution Approach 1:
The patent divides backup policies into stage-specific sub-policies that can be independently configured and optimized. Each stage (collection, transfer, restoration) has its own policy parameters, allowing administrators to manage complexity through modular policy structures while achieving optimal performance for each phase.
Data Source
AI summary
Optimizing multiple backup operations for multiple sources by discovering shared data routes between sources and targets for backup operations performed in a certain time period. Backups using shared routes use a Best Time Algorithm that determines a size of a dataset to be saved an available bandwidth in each transfer window of the time period, and then determines an order of the backup operations based on first ordering the backups based on decreasing dataset size and second ordering the transfer windows in order of decreasing bandwidth, and matching the backups to the transfer windows in accordance with the first ordering and second ordering. The optimum time represents a time to initiate each backup of the multiple backups that results in a shortest data transfer time over the entire time period.