Adaptive Health Driven Network Slicing for Data Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data migration processes in computing systems often fail to prioritize urgent data migrations, leading to performance degradation in production workloads due to inadequate handling of predicted failure states in storage and processing components.
Innovation Solution
The implementation of network slicing to prioritize urgent data migration by determining confidence levels of predicted failure states in storage and processing components, mapping data to high-priority network slices, and reordering data for efficient migration, ensuring that urgent data is migrated effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data migration is performed without prioritization, then all data is migrated, but urgent data migration is delayed causing performance degradation
Solution Approach 1:
The patent segments the network into multiple network slices with different priority levels. Urgent data migrations are assigned to high-priority slices while non-urgent migrations use lower-priority slices, enabling differentiated handling that prevents performance degradation in production workloads while maintaining overall migration productivity.
Solution Approach 2:
The patent changes the parameter of network resource allocation by dynamically assigning different priority levels to data migration tasks based on urgency. This parameter change enables the system to prioritize urgent migrations without sacrificing overall migration throughput, resolving the contradiction between migration speed and production performance.
2Reliability
If network slicing is implemented to prioritize urgent data, then urgent data migration is improved, but system complexity increases
Solution Approach 1:
The patent implements a universal network slicing framework that can handle multiple types of data migrations with different priorities using the same infrastructure. This multi-functionality approach ensures urgent data migration priority while avoiding excessive complexity by reusing existing network resources across different slice types.
Solution Approach 2:
The patent introduces a network slice manager as an intermediary component that handles the complexity of slice creation, assignment, and management. This mediator abstracts the complex network slicing operations from the data migration processes, ensuring urgent data priority while containing system complexity within the management layer.
3Productivity
If data is reordered for efficient migration, then migration efficiency is improved, but processing time increases
Solution Approach 1:
The patent performs data reordering as a preliminary action before initiating the actual data migration process. By pre-ordering data according to priority and destination, the system eliminates the need for complex reordering during migration, improving overall migration efficiency while accepting a controlled amount of preprocessing time.
Solution Approach 2:
The patent implements periodic reordering of data batches during the migration process rather than reordering all data at once. This periodic approach maintains migration efficiency by keeping data flows moving while periodically adjusting order, reducing the total time lost to reordering operations.
Data Source
AI summary
A method for managing data migration includes: obtaining a confidence level of a predicted failure state of a storage component (SC) of a source device; making a first determination that the confidence level exceeds a minimum threshold; making, based on the first determination, a second determination that a high priority network slice of a network exists; mapping data stored in the SC to the high priority network slice; migrating the data to a target device; after migrating the data to the target device: obtaining a confidence level of a predicted failure state of a processing component (PC); making a third determination that the confidence level of the predicted failure state of the PC exceeds a maximum threshold; mapping, based on the third determination, non-migrated data stored in the source device to the high priority network slice; and migrating the non-migrated data to the target device.