I/O Scheduling Measurement in Distributed Virtual Infrastructure
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In distributed virtual infrastructure computing systems, measuring input/output (I/O) scheduling characteristics is challenging due to the complexity of managing tens to hundreds of physical processing elements, which can impact compliance with service level agreements (SLAs) between infrastructure providers and tenants.
Innovation Solution
A method is introduced to measure I/O scheduling characteristics by deploying virtual machines under a hypervisor, executing test workloads, and collecting timing information to identify characteristics of the I/O schedule, using different workload modes (sequential, burst, and random) to explore time and space locality of I/O schedulers, and employing classification rules for open-source and close-source hypervisors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple virtual machines are deployed on tens to hundreds of physical processing elements, then the distributed virtual infrastructure provides enhanced computing capacity and resource sharing, but I/O scheduling becomes increasingly complex and difficult to manage
Solution Approach 1:
The patent introduces a measurement system that acts as an intermediary between the hypervisor and the I/O operations. This system includes probes deployed in virtual machines that collect timing information about I/O commands, and an analysis system that processes this data to characterize hypervisor I/O scheduling behavior. This intermediary measurement layer simplifies the management complexity by providing automated characterization without requiring manual configuration or intervention in the complex I/O scheduling paths.
Solution Approach 2:
The patent implements a feedback mechanism where timing information from I/O commands is collected, analyzed, and used to characterize hypervisor scheduling behavior. The system continuously monitors I/O command timing data, analyzes patterns to identify scheduling characteristics, and uses this feedback to confirm SLA compliance. This closed-loop feedback approach automates the management of I/O scheduling complexity through continuous observation and adaptive characterization.
2Ease of manufacture
If traditional I/O scheduling measurement methods are used, then the implementation is simpler, but they cannot accurately capture the scheduling characteristics in distributed virtual infrastructure with tens to hundreds of physical processing elements
Solution Approach 1:
The patent segments the I/O scheduling measurement problem into distinct components: probes deployed in individual virtual machines that collect local timing information, a measurement system that aggregates data from multiple probes, and an analysis system that characterizes hypervisor behavior. This segmentation allows accurate measurement across distributed virtual infrastructure by breaking down the complex global measurement task into manageable local measurements that are then synthesized.
Solution Approach 2:
The patent adds a new dimension to I/O scheduling measurement by introducing time-based timing information collection. Instead of merely counting I/O operations, the system measures the timing of I/O command submission and completion across multiple virtual machines and physical processing elements. This temporal dimension enables accurate characterization of scheduling latency and behavior patterns that traditional methods cannot capture.
3Measurement precision
If comprehensive timing information is collected from all virtual machines, then accurate I/O scheduling characteristics can be identified, but the data collection and processing overhead increases
Solution Approach 1:
The patent applies partial action by collecting timing information selectively rather than comprehensively from all possible sources. The measurement system focuses on collecting timing data for I/O commands that are relevant to characterizing hypervisor scheduling behavior, rather than attempting to capture every I/O operation in the system. This selective collection reduces processing overhead while maintaining sufficient accuracy for SLA compliance verification.
Data Source
AI summary
Techniques for measuring input/output (I/O) scheduling characteristics in a computing system implemented with distributed virtual infrastructure. In one example, a method comprises the following steps. A plurality of virtual machines is deployed in a computing system implemented in a distributed virtual infrastructure managed by at least one hypervisor component. At least one test workload is executed on the plurality of virtual machines in accordance with the at least one hypervisor component. Timing information is collected from the plurality of virtual machines during execution of the at least one test workload. Based on at least a portion of the collected timing information, one or more characteristics are identified of an input/output schedule employed by the at least one hypervisor component during execution of the at least one test workload on the plurality of virtual machines.