Intermediary Storage for Virtual Machine Migration Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Live virtual machine migration between hosts can fail due to destination host or network issues, leading to significant downtime, especially during geographically remote migrations, and result in data loss or degradation due to page-faults and network-faults during suspended states.
Innovation Solution
The implementation of an intermediary storage device that synchronizes and stores virtual machine memory pages during the warm-up phase, allowing for seamless recovery in case of failures by providing the memory state from the intermediary device to the destination host or an alternative host, thereby reducing downtime and enhancing migration reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If live migration is performed without an intermediary storage device, then device complexity is reduced, but migration reliability deteriorates due to potential failures during memory copying
Solution Approach 1:
An intermediary storage device is introduced between the source host and destination host to store memory pages during migration. This mediator ensures that even if the destination host or network fails, the memory state can be recovered from the intermediary device, significantly improving migration reliability without requiring complex backup mechanisms at the host level
2Productivity
If memory copying is performed during suspended state, then migration speed is improved, but loss of information increases due to page-faults and network-faults
Solution Approach 1:
Memory pages are copied to the intermediary storage device during the warm-up phase while the VM is still running, before suspension occurs. This preliminary action ensures that the memory state is already prepared and stored safely, allowing the VM to be suspended without risking data loss from page-faults or network failures during the actual migration completion
3Manufacturing precision
If the VM is suspended during migration, then migration precision is improved by copying memory state, but loss of time increases due to suspension downtime
Solution Approach 1:
The memory copying process is initiated during the warm-up phase while the VM is running, before suspension is required. By preparing the memory state in advance and storing it in the intermediary device, the actual suspension and completion phases can proceed quickly, minimizing the total downtime while maintaining migration precision
Data Source
AI summary
Systems and methods for live migration of a virtual machine are provided. A migration manager can send a request to a source host machine to migrate a virtual machine to a first destination host machine. The migration manager can determine that the migration of the virtual machine from the source host machine to the first destination host machine has failed. The migration manager can identify an intermediary storage device used during the migration to store a state of the virtual machine. The migration manager identifies a second destination host machine for the virtual machine migration. The migration manager causes the second destination host machine to obtain the state of the virtual machine from the intermediary storage device to migrate the virtual machine to the second destination host machine.