Storage Management Device Optimizes Volume Placement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Distributed storage systems face challenges in efficiently eliminating imbalances in storage service load and capacity among storage apparatuses, as existing methods either overlook volume size during initial placement or take too long to balance large data migrations.
Innovation Solution
A computer system with a storage management device that selects storage apparatuses based on feature amounts to create new volumes, optimizing volume placement and migration to reduce imbalance and shorten migration time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If volume creation is performed without considering volume size during initial placement, then the placement process is simple, but the imbalance in storage service load and capacity cannot be sufficiently improved
Solution Approach 1:
The patent introduces feature amounts as parameters for volume placement decisions. When creating or migrating volumes, the storage management device calculates feature amounts based on volume size, storage service load, and usage capacity, then selects target storage apparatuses that optimize these parameters to reduce imbalance while maintaining manageable placement complexity.
2Reliability
If large data volumes are migrated to eliminate imbalance, then the load balancing effectiveness is improved, but the migration time becomes excessively long
Solution Approach 1:
The patent calculates feature amounts that combine volume size, storage service load, and usage capacity to identify optimal migration targets. By selecting volumes with appropriate feature amounts, the system achieves effective load balancing while avoiding excessively long migration times that would result from moving only the smallest volumes.
Solution Approach 2:
The storage management device continuously monitors storage service load and usage capacity, calculates feature amounts based on this feedback, and dynamically selects migration targets. This feedback mechanism ensures that migration decisions are made based on current system state, optimizing both load balancing effectiveness and migration time.
3Loss of time
If only the smallest data volumes are migrated to reduce imbalance, then the migration time is shortened, but the imbalance elimination is insufficient
Solution Approach 1:
The patent uses feature amounts that incorporate volume size, storage service load, and usage capacity together. This multi-parameter approach allows the system to select volumes for migration that are not necessarily the smallest but still achieve effective imbalance reduction, balancing migration time with elimination effectiveness.
Data Source
AI summary
To more effectively eliminate a resource imbalance among storage apparatuses, and shorten the time required for elimination in a computer system including a plurality of storage apparatuses and a computer. In a computer system including a plurality of storage apparatuses that provide a volume to a computer and a storage management device that manages the plurality of storage apparatuses, in a case where there is an instruction to create a new volume, the storage management device compares a distribution of feature amounts of all volumes provided by each storage apparatus and a distribution of feature amounts of all volumes in a case of providing a newly created volume in each storage apparatus, and instructs a storage apparatus having a largest difference in the distributions of the feature amounts to create a volume.