Storage Server I/O Priority Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage server management systems are inefficient as they throttle lower priority IO requests in favor of high priority requests, leading to system overload and suboptimal performance due to static mapping of storage requirements and service classes, resulting in minimal fulfillment of low priority requests.
Innovation Solution
Dynamic priority assignment to IO requests based on performance levels and importance, with throttling of lower priority requests when high priority requests do not meet their targets, allowing resource reallocation to enhance high priority request performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If lower priority IO requests are throttled based on static mapping between storage requirements and storage service classes, then high priority IO requests can be fulfilled, but the system becomes overloaded by the volume of high priority IO requests in a queue and low priority requests are minimally fulfilled
Solution Approach 1:
The patent implements dynamic priority assignment for IO requests, where priorities are not fixed but adjusted based on current system conditions, performance targets, and queue states. This allows the system to adaptively balance between high and low priority requests, preventing overload while maintaining quality of service guarantees.
Solution Approach 2:
The system changes the parameter of request priority dynamically based on performance levels and target comparisons. When high priority requests fail to meet performance targets, the system adjusts priorities to allow low priority requests to proceed, thereby optimizing overall system throughput and preventing queue overload.
2Device complexity
If static mapping between storage requirements and storage service classes is used, then resource allocation is simplified, but system operation efficiency decreases and low priority requests are minimally fulfilled
Solution Approach 1:
The patent replaces static mapping with dynamic priority assignment that responds to real-time performance conditions. This dynamic approach optimizes system operation efficiency by adjusting resource allocation based on actual performance targets and system state, while maintaining manageable complexity through automated decision-making rules.
Solution Approach 2:
The system incorporates feedback loops that monitor performance levels of high priority requests and use this information to dynamically adjust priority assignments. This feedback mechanism enables efficient system operation by automatically reallocating resources when performance targets are not met, without requiring complex manual intervention.
Data Source
AI summary
For managing a storage server having improving overall system performance, a first input/output (I/O) request is received. A first priority level is dynamically assigned to the first I/O request, the first I/O request associated with a performance level for an application residing on a host in communication with the storage server. A second I/O request of a second priority level is throttled to allow at least a portion of a predetermined amount of resources previously designated for performing the second I/O request to be re-allocated to performing the first I/O request. The second priority level is different than the first priority level.