Container Application Protocol Switching for Storage Performance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques face challenges in dynamically switching host devices between different storage access protocols, such as SCSI and NVMe, due to issues like congestion and performance conditions, which can degrade system performance and complicate migration processes.
Innovation Solution
Implementing a multi-path layer with a multi-path input-output (MPIO) driver that monitors performance across multiple paths and dynamically moves applications between containers using different access protocols, such as SCSI-FC and NVMeF, based on detected performance conditions to optimize access and improve overall system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a host device is configured to utilize only one access protocol at a time to access a particular logical storage device, then the configuration is simple and stable, but the system cannot adapt to performance conditions and congestion, resulting in suboptimal IO processing performance
Solution Approach 1:
The system dynamically switches between different access protocols (SCSI and NVMe) based on real-time performance monitoring. The host device can transition from a static single-protocol configuration to a dynamic multi-protocol configuration where the optimal protocol is selected automatically based on congestion conditions, IO patterns, and performance metrics, thereby resolving the contradiction between adaptability and complexity.
Solution Approach 2:
The invention changes the operational parameters of the host device by allowing it to switch between different access protocols. The system monitors performance parameters such as IO latency, throughput, and congestion levels, and adjusts the access protocol parameter accordingly. This parameter change enables the system to adapt to varying performance conditions while maintaining manageable configuration complexity through automated decision-making.
2Reliability
If conventional techniques are used to transform host device access from SCSI to NVMe, then the migration process is straightforward, but system performance degrades due to congestion and errors on existing paths
Solution Approach 1:
The system implements continuous performance monitoring and feedback mechanisms that track IO operations, congestion levels, and errors for each access protocol path. Based on this feedback, the system automatically determines when to switch protocols to maintain application survivability and optimize IO processing performance, preventing degradation caused by congestion on existing paths.
Solution Approach 2:
The invention segments the storage access paths into distinct protocol-specific paths (SCSI paths and NVMe paths) with independent performance monitoring and selection. This segmentation allows the system to evaluate each path's performance independently and switch between them based on real-time conditions, thereby maintaining reliability while optimizing productivity without being constrained by a single path's congestion or error conditions.
3Adaptability or versatility
If multiple access protocols are supported by the storage system, then protocol versatility is improved, but reconfiguring the host device to utilize a different access protocol becomes problematic and complex
Solution Approach 1:
The system enables self-service protocol selection where the host device automatically chooses the optimal access protocol based on monitored performance conditions. The reconfiguration process is automated through the multi-path layer and performance monitoring mechanisms, eliminating the need for manual administrator intervention. This self-service approach maintains protocol versatility while significantly improving the ease of reconfiguration by removing complex manual setup requirements.
Data Source
AI summary
An apparatus in one embodiment comprises at least one processing device. The at least one processing device is configured to monitor performance of respective ones of a plurality of paths for accessing a logical storage device, and responsive to detection of at least one specified condition in the monitored performance relating to at least a subset of the paths, to move at least one application from a first container that utilizes a first access protocol to access the logical storage device to a second container that utilizes a second access protocol different than the first access protocol to access the logical storage device. For example, in some embodiments, the at least one processing device is configured to move an application from first container that utilizes a SCSI access protocol to a second container that utilizes an NVMe access protocol, and vice versa, responsive to detected performance issues.