Network Card Shutdown Interface for Ordered VM Power-Failure Shutdown
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information processing systems face difficulties in properly shutting down virtual machines, particularly during power failures, without consuming resources of the host devices.
Innovation Solution
A method involving a controller application on a network card that orchestrates the shutdown of user-created virtual machines, followed by system-created virtual machines and host devices, using a shutdown interface to ensure orderly shutdown without relying on host device resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If virtual machines are shut down during power failure without a coordinated interface, then the shutdown process is simple, but metadata corruption and configuration errors occur
Solution Approach 1:
A shutdown interface is introduced as an intermediary component between the power failure detection mechanism and the virtual machine shutdown process. This interface coordinates the shutdown sequence, ensuring user-created virtual machines are stopped before system-created ones, thereby preventing metadata corruption and configuration errors while maintaining system reliability.
Solution Approach 2:
The shutdown interface performs preliminary actions by first shutting down user-created virtual machines before proceeding to system-created virtual machines. This staged approach ensures that critical user data is preserved and metadata is not corrupted during the shutdown process, addressing the reliability issue.
2Productivity
If host device resources are used for shutdown operations, then shutdown can be performed, but resource consumption increases during critical power failure
Solution Approach 1:
The network card performs shutdown operations autonomously without requiring host device resources. The shutdown interface on the network card directly manages the shutdown process of virtual machines, enabling the system to serve itself during power failure conditions and avoiding additional resource consumption from the host device.
3Ease of operation
If system-created virtual machines are shut down first, then shutdown process is simplified, but metadata corruption and configuration errors occur
Solution Approach 1:
The shutdown interface implements preliminary action by first shutting down user-created virtual machines before system-created virtual machines. This sequence preserves metadata and configuration integrity because user data is evacuated first, while system-created virtual machines are shut down afterward in a controlled manner, preventing corruption.
Solution Approach 2:
The shutdown process is segmented into distinct phases: first shutting down user-created virtual machines, then shutting down system-created virtual machines. This segmentation allows each phase to be managed independently, ensuring data integrity while maintaining operational simplicity through automated coordination.
Data Source
AI summary
Techniques are provided for orchestrating a shutdown of virtual machines using a shutdown interface and a network card. One method comprises receiving a message sent by a controller application, executing on a network card of a hyperconverged appliance, wherein the message is sent by the controller application after shutting down one or more user-created virtual machines in response to a detected power failure of the hyperconverged appliance; performing the following steps, by a shutdown interface, in response to receiving the message: shutting down one or more system-created virtual machines and one or more physical host devices of the hyperconverged appliance. The controller application may pass control to the shutdown interface and may shut down in response to the completion of the shutting down of the user-created virtual machines. The execution of the controller application on the network card may not employ resources of the physical host devices.