Virtual Storage Migration Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems face challenges in migrating virtual storage systems from a source to a destination without disrupting client access, as they struggle to predict the likelihood of successful migration and ensure the destination storage system can handle post-migration workload efficiently.
Innovation Solution
A processor-executable management application estimates the likelihood of migration success by analyzing workload and utilization of both source and destination storage systems before the cut-over duration, comparing these estimates to threshold values to decide whether to proceed or abort the migration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If migration operation proceeds during cut-over duration, then migration completion is achieved, but client access to storage system is restricted
Solution Approach 1:
The system performs preliminary actions before the cut-over duration by copying data from source to destination storage during non-cut-over periods, preparing the migration in advance so that the actual cut-over can be completed quickly with minimal client access disruption
Solution Approach 2:
The system maintains continuity of useful action by allowing client access to continue during data copying phases and by using predictive analytics to ensure migration completion before cut-over, thereby minimizing interruption to storage operations
2Productivity
If migration operation is performed without prediction, then migration speed is improved, but migration reliability deteriorates
Solution Approach 1:
The system performs preliminary predictive analytics before executing the migration operation, estimating workload and likelihood of success in advance to ensure reliable migration completion without sacrificing overall migration speed
Solution Approach 2:
The system uses feedback from predictive analytics about workload estimation and success probability to dynamically adjust migration timing and resource allocation, ensuring both high migration speed and reliable completion
3Reliability
If workload estimation is performed before cut-over, then migration reliability is improved, but processing time increases
Solution Approach 1:
The system performs partial workload estimation focusing only on critical parameters needed for success prediction rather than comprehensive analysis, achieving sufficient reliability improvement with minimal time loss
Solution Approach 2:
The system uses existing workload data and performance metrics that are already being collected for other purposes, performing self-service estimation without requiring additional dedicated analysis time
Data Source
AI summary
Method and system for migrating a virtual storage system from a source storage system having access to a source storage device to a destination storage system having access to a destination storage device is provided. A processor executable management application estimates a likelihood of success for a migration operation before the migration operation enters a cut-over duration during which client access to the source storage system and the destination storage system is restricted. The migration operation enters the cut-over duration if there is high likelihood of success for completing the migration during the cut-over duration or aborted, if there is a low likelihood of success for completing the migration during the cut-over duration.