Data Service-Aware IO Scheduler Request Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current IO schedulers in computing environments are unaware of data service segment requirements, leading to significant request misalignment and unnecessary processing overhead, especially in multi-core, multi-thread host environments, which affects the performance of deduplication and compression processes.
Innovation Solution
Implementing a data service-aware IO scheduler that aligns requests with the underlying data services' segment size requirements through a two-phase methodology involving local and global queues, with configuration modules that set preferred segment sizes and optimize merge/sort operations to ensure alignment and reduce latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If IO scheduler performs merging and sorting operations on pending IO requests, then disk seek time is reduced and system throughput is improved, but request alignment with data service segment size is lost and processing overhead increases
Solution Approach 1:
The patent segments the IO request processing into distinct phases: a first phase for merging/sorting operations to improve throughput, and a second phase for alignment operations to ensure segment size compatibility. This segmentation allows each phase to optimize for its specific goal without interfering with the other.
Solution Approach 2:
The patent performs merging and sorting operations as preliminary actions before the final alignment and dispatch to data services. By completing these optimization operations first, the system prepares requests for efficient processing while maintaining the ability to align them subsequently with segment size requirements.
2Manufacturing precision
If IO scheduler issues requests in received order without merging or sorting, then request alignment is maintained, but disk seek time increases and system throughput decreases
Solution Approach 1:
The patent divides the request processing workflow into distinct segments: an optimization segment that performs merging/sorting, and an alignment segment that ensures segment size compatibility before dispatch. This allows the system to benefit from both optimization and alignment.
Solution Approach 2:
The patent introduces an intermediary alignment operation between the merging/sorting phase and the final dispatch to data services. This intermediary ensures that requests maintain both optimization benefits and segment size alignment, acting as a bridge between the two requirements.
3Speed
If IO scheduler performs extensive merging and sorting operations, then disk seek efficiency is improved, but processing latency increases
Solution Approach 1:
The patent applies partial merging and sorting operations rather than exhaustive operations. It performs enough merging/sorting to achieve meaningful disk seek optimization while setting limits to prevent excessive processing latency, balancing efficiency gains with time constraints.
Solution Approach 2:
The patent segments the merging/sorting operations into controlled phases with defined boundaries, preventing excessive processing time by limiting the scope and duration of optimization operations while still achieving meaningful disk seek efficiency improvements.
Data Source
AI summary
A method of request scheduling in a computing environment comprises the following steps. One or more requests to at least one of read data from and write data to one or more storage devices in the computing environment are obtained from a host device. The one or more requests are aligned corresponding to a segment size for which one or more data services in the computing environment are configured to process data. The one or more aligned requests are dispatched to the one or more data services prior to sending the one or more requests to the one or more storage devices.