Hypervisor VM Reset via Direct Agent Error Messages
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual machine management systems face delays and errors due to uncertainty in the cause of missed heartbeat signals, reliance on external management servers, and increased complexity in initiating corrective actions, which can lead to unnecessary resets or slow recovery.
Innovation Solution
A hypervisor receives error messages directly from virtual machines via trusted communication channels, allowing it to initiate corrective actions such as resets or reconfigurations without delays, and determines dependencies between virtual machines to address issues effectively, thereby reducing reliance on external components and improving response times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the hypervisor waits for a period of time before resetting the virtual machine in response to a missed heartbeat, then unnecessary resets are prevented, but recovery time is increased when a reset is actually needed
Solution Approach 1:
The hypervisor performs preliminary actions by sending diagnostic commands to the virtual machine agent before initiating a reset. This allows the system to gather information about the virtual machine's state and determine whether a reset is actually necessary, thereby avoiding unnecessary resets while enabling faster recovery when problems do exist.
Solution Approach 2:
The system implements feedback mechanisms where the virtual machine agent responds to hypervisor commands with status information. This feedback loop allows the hypervisor to make informed decisions about whether to reset the virtual machine, balancing the prevention of unnecessary resets with timely recovery when needed.
2Ease of operation
If the hypervisor uses an external management server to initiate corrective actions, then centralized control is achieved, but delays and connection errors increase
Solution Approach 1:
The patent extracts the corrective action initiation capability from the external management server and places it directly in the hypervisor. This allows the hypervisor to independently send commands to virtual machine agents and initiate resets without waiting for management server responses, significantly reducing response time while maintaining centralized management for policy and coordination.
Solution Approach 2:
The virtual machine agent acts as an intermediary between the hypervisor and the virtual machine operating system. The agent receives commands from the hypervisor and executes them directly, eliminating the need for management server involvement in the critical path of corrective actions while preserving centralized management capabilities for oversight and policy enforcement.
3Ease of operation
If the hypervisor initiates corrective actions through the virtual machine's operating system, then user control is maintained, but the critical path is extended and recovery is slowed
Solution Approach 1:
The patent extracts the corrective action execution path from the virtual machine operating system and creates a separate channel through the virtual machine agent. This allows the hypervisor to initiate and complete corrective actions directly without traversing the operating system's command interface, reducing critical path complexity and enabling faster recovery while maintaining user control through hypervisor-level authorization.
Data Source
AI summary
Exemplary methods, apparatuses, and systems include a hypervisor receiving an error message from an agent within a first virtual machine run by the hypervisor. In response to the error message, the hypervisor determines and initiates a corrective action for the hypervisor to take in response to the error message. An exemplary corrective action includes initiating a reset of the first virtual machine or a reset of a second virtual machine.