Hypervisor Virtual Machine Migration Coordination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The migration of virtual machine instances between host computer systems is often cumbersome, particularly for large or complex systems, as existing methods require halting the VM, copying its state, and restarting, which can be time-consuming and disrupt user interaction, and may result in operational issues due to volatile system resources.
Innovation Solution
A live migration approach is implemented, where the VM is paused and resumed on a new host, with iterative copying of memory and state, and resource management phases to minimize downtime and ensure resource consistency, using a hypervisor to coordinate the migration through phases like pause, route, standby lease, packet forward, and active lease stages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the virtual computer system is halted and memory/state is copied to a different host, then the system can be migrated to avoid resource scarcity, but the migration process takes a significant amount of time and user interaction is eliminated or severely restricted
Solution Approach 1:
The migration process is divided into distinct phases: initial state copy, iterative memory state copying, and final state synchronization. This segmentation allows the system to maintain operational continuity while progressively transferring state, reducing overall migration time and maintaining user interaction capability throughout most of the process.
Solution Approach 2:
The system performs preliminary copying of memory and state information before the actual migration is complete. By pre-copying state data and maintaining the ability to roll back if issues arise, the system reduces the time required for final migration execution while ensuring reliability.
2Reliability
If the virtual computer system is halted for migration, then system state can be copied, but the ability of a user to interact with the virtual computer system is eliminated or at least severely restricted
Solution Approach 1:
The system performs preliminary copying of memory and state information before the actual migration is complete. By pre-copying state data and maintaining the ability to roll back if issues arise, the system reduces the time required for final migration execution while ensuring reliability.
Solution Approach 2:
The system maintains continuous user interaction capability throughout the migration process by allowing the virtual computer system to remain operational during state copying. The migration occurs in the background with minimal disruption, and the system can seamlessly switch to the new host without interrupting user workflows.
3Adaptability or versatility
If system resources such as attached storage and network connections are accessed during migration, then the migrated virtual computer system may differ significantly from the original, but this introduces operational issues
Solution Approach 1:
The system employs feedback mechanisms to monitor resource states during migration and make adjustments to maintain consistency. By continuously comparing the original and migrated system states and applying corrective actions, the system ensures that resource attachments remain consistent across hosts.
Solution Approach 2:
The system dynamically adjusts resource parameters during migration to maintain operational consistency. By monitoring and modifying resource states in response to changing conditions, the system ensures that storage and network connections remain stable and consistent throughout the migration process.
Data Source
AI summary
Techniques for managing a critical phase of a virtual machine migration are described herein. During the critical phase, which must be kept as short as possible, a virtual machine instance is paused, network packets are rerouted to a temporary packet queue, and messages associated with managing synchronization of block storage devices are dispatched to an entity configured to manage those connections. After the block storage devices are synchronized, the network packets are released from the temporary packet queue to the new location of the virtual machine instance.