Virtual Machine Migration Control via Dynamic Pattern Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In server virtualization environments, preset control patterns for dynamic changes and migrations of virtual computers are not effective, leading to prolonged migration times when multiple virtual computers need to be moved, making immediate migration challenging.
Innovation Solution
A method involving a management server that detects triggers, acquires relationships between virtualization units and virtual machines, specifies migration targets, determines migration destinations, transforms control patterns, generates and evaluates migration plans, and selects the most suitable plan for immediate virtual machine migration based on dynamic information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple virtual computers are migrated individually according to preset control patterns, then migration can be performed systematically, but the time required for migration increases significantly
Solution Approach 1:
The patent combines multiple individual virtual machine migration operations into a single LPAR-level migration operation. Instead of migrating each virtual machine separately according to preset control patterns, the system identifies that multiple VMs share common migration characteristics and consolidates their migration into one batch operation at the LPAR level, thereby significantly reducing total migration time while maintaining systematic control.
Solution Approach 2:
The patent introduces dynamic control pattern transformation where the management server transforms control patterns based on real-time system state information. The system dynamically determines migration strategies by evaluating current LPAR and VM states, transforming static preset patterns into adaptive dynamic patterns that optimize migration timing and sequencing, thus reducing overall migration time while ensuring reliability.
2Extent of automation
If preset control patterns are used for virtual machine migration, then automated control is achieved, but the patterns cannot adapt to dynamic system situations
Solution Approach 1:
The patent implements feedback mechanisms where the management server continuously acquires system state information including LPAR status, VM configurations, and resource utilization data. This feedback is used to dynamically transform control patterns based on actual system conditions. The system compares current states with target states defined in control patterns and automatically adjusts migration timing and sequencing, enabling automated control that adapts to dynamic situations.
Solution Approach 2:
The patent changes parameters of control patterns dynamically based on system state. The management server transforms control patterns by modifying parameters such as migration timing, sequencing, and resource allocation based on real-time acquisition of LPAR and VM state information. This allows preset automated patterns to adapt to varying system conditions without manual intervention, maintaining both automation and adaptability.
3Ease of operation
If individual virtual computers are migrated separately, then precise control over each VM is achieved, but migration efficiency decreases
Solution Approach 1:
The patent segments the control granularity from individual VM level to LPAR level. Instead of controlling and migrating each virtual machine independently with full precision, the system segments the management scope to the LPAR level, where multiple VMs are managed as a unified unit. This segmentation maintains operational control precision at the LPAR level while improving migration efficiency by handling multiple VMs simultaneously through batch operations.
Solution Approach 2:
The patent applies universality by creating a unified LPAR-level migration mechanism that serves multiple virtual machines simultaneously. The management server uses a single LPAR migration operation to migrate multiple VMs that share common characteristics, making the migration process multi-functional and efficient. This approach maintains control precision through unified management while significantly improving productivity by reducing the number of separate migration operations required.
Data Source
AI summary
A method of controlling a virtual machine, including: detecting a predetermined trigger by a management server; acquiring a relationship between a virtualization unit and a virtual machine operated on a physical computer; specifying a virtual machine as a migration target based on preset trigger definition information; determining a physical computer of a migration destination based on a preset first control pattern to generate a first plan to migrate the virtual machine as the migration destination; transforming the first control pattern to a second control pattern; generating a second plan to migrate the virtual machine as the migration destination; performing arithmetic an evaluation value of the first plan and an evaluation value of the second plan; selecting one of the evaluation values meeting a predetermined condition as an execution plan; and migrating by the management server the virtual machine as the migration target based on the selected execution plan.