Hypervisor Execution Servers for Multitasking Guest Temporal Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for hosting multitasking guests on host systems fail to effectively manage time-driven and event-driven behaviors, leading to potential interrupt latency and lack of temporal isolation, especially when handling faulty guests.
Innovation Solution
A server-based approach is employed, utilizing deferrable and periodic execution servers to allocate CPU capacity, with watchdog monitoring to ensure timely completion of event-driven tasks and separate priority levels for hypervisor tasks, ensuring minimal interrupt latency and temporal isolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a hypervisor hosts multiple virtual machines on a host system, then resource utilization and system functionality are improved, but interrupt latency and temporal isolation deteriorate when handling event-driven tasks
Solution Approach 1:
The patent segments the virtual machine execution environment into multiple execution servers (deferrable execution server and periodic execution server), each responsible for specific task types. This segmentation allows event-driven tasks to be handled by the deferrable server with guaranteed interrupt response, while time-driven tasks are managed by the periodic server, thus resolving the contradiction between hosting multiple VMs and maintaining low interrupt latency.
Solution Approach 2:
The patent introduces execution servers as intermediary layers between the hypervisor and virtual machines. These servers act as mediators that manage task scheduling and execution, providing temporal isolation and guaranteed response times for event-driven tasks while enabling the host system to efficiently host multiple VMs with different timing requirements.
2Speed
If execution capacity is allocated to handle event-driven tasks with minimal interrupt latency, then real-time performance is improved, but temporal isolation and reliability deteriorate when faulty guests fail to signal completion
Solution Approach 1:
The patent implements feedback mechanisms through watchdog timers that monitor the execution of tasks in virtual machines. When a task is dispatched, the watchdog timer tracks its execution and expects a completion signal. If the task fails to signal completion within the expected time, the watchdog detects the fault and triggers appropriate error handling, thus maintaining temporal isolation and reliability while preserving fast interrupt response through the deferrable execution server.
Solution Approach 2:
The patent establishes preliminary monitoring structures (watchdog timers) before tasks are executed. These watchdogs are pre-configured with timeout values and are automatically activated when tasks are dispatched. This preliminary action ensures that if a faulty guest fails to signal completion, the system can detect and respond to the error, maintaining reliability without compromising the real-time performance achieved through the deferrable execution server.
3Reliability
If separate priority levels are implemented for hypervisor tasks, then temporal isolation is improved, but system complexity increases
Solution Approach 1:
The patent segments the scheduling system into distinct priority levels and execution servers with clearly defined responsibilities. The deferrable execution server handles event-driven tasks at high priority for minimal latency, while the periodic execution server manages time-driven tasks. This segmentation provides temporal isolation through structured priority management without creating excessive complexity, as each segment has well-defined rules and boundaries.
Data Source
AI summary
Method for hosting a multitasking guest on a host system, wherein the guest is allocated a periodic execution server and a deferrable execution server associated with a partition of the system, event-driven tasks of the guest are assigned to the deferrable execution server, time-driven tasks of the guest are assigned to the periodic execution server, a hypervisor of the execution servers is assigned to a first priority group of the system, the deferrable execution server is assigned to a second priority group of the system, the periodic execution server is assigned to a third priority group of the system, and a preemptive task scheduler maintains a descending priority order among the priority groups while the execution servers execute the tasks concurrently.