SAN Multipath Failover via Non-Overlapping Endpoint Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional multipathing solutions in storage area networks (SANs) randomly select alternative paths during failover, leading to potential delays and increased likelihood of choosing a non-functional endpoint, resulting in I/O request timeouts and reduced system availability.
Innovation Solution
A methodology that identifies a failover path with endpoints different from the suspect or non-functional path, utilizing SAN topology information and subpath failover groups to select an alternative path that avoids the disabled endpoint, ensuring a higher probability of successful failover without needing to identify the failed endpoint.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional multipathing solutions randomly choose alternative paths during failover, then the failover process is simple to implement, but the likelihood of choosing a non-functional endpoint increases, leading to I/O request timeouts and reduced system availability
Solution Approach 1:
The system performs preliminary actions by pre-establishing multiple paths with different endpoints before a failure occurs. When a failure is detected, the multipathing solution can immediately switch to an alternative path without needing to analyze the current state, because the alternative paths are already prepared with known non-overlapping endpoints. This preliminary preparation resolves the contradiction by ensuring reliability through pre-planned failover paths while avoiding the complexity of real-time path analysis.
Solution Approach 2:
The system implements feedback mechanisms to monitor the operational status of paths and endpoints. When a failure is detected in one path, the feedback information triggers the selection of an alternative path with non-overlapping endpoints. This feedback loop ensures that the multipathing solution learns from failures and selects safer alternative paths, improving system availability while maintaining manageable complexity through automated feedback-driven decisions.
2Loss of time
If the multipathing solution quickly fails over to an alternative path, then I/O request timeouts are reduced, but the complexity of path selection increases to ensure the alternative path is functional
Solution Approach 1:
The system segments paths into distinct groups based on their endpoints, creating independent failover groups where paths share common endpoints. This segmentation allows the multipathing solution to quickly identify and switch between groups without analyzing individual path status in detail. When a failure occurs in one segment, the system can immediately select from other segments with non-overlapping endpoints, achieving fast failover with reduced complexity through structured path organization.
Solution Approach 2:
The system applies local quality by making path selection decisions based on the specific characteristics of individual paths and endpoints. Instead of treating all paths uniformly, the multipathing solution evaluates the local properties of paths (such as endpoint configurations) to make optimized failover decisions. This localized approach enables quick and intelligent path selection by focusing on the specific qualities of candidate paths rather than performing comprehensive global analysis.
3Reliability
If conventional multipathing solutions use random path selection, then the implementation is straightforward, but the probability of selecting a successful alternative path decreases
Solution Approach 1:
The system performs preliminary organization of paths into failover groups with non-overlapping endpoints before failures occur. This pre-arrangement ensures that when a failure happens, the multipathing solution can immediately select a successful alternative path from the pre-prepared groups, dramatically increasing failover success probability. The preliminary structuring of paths eliminates the need for complex real-time analysis, maintaining implementation simplicity while achieving high reliability through pre-planned path configurations.
Data Source
AI summary
A first path for forwarding an I/O request from a host device to a disk in a disk array is identified. The first path includes two endpoints (a first initiator endpoint on the host device and a first target endpoint on the disk array) separated by a storage area network. In response to an indication that the first path is non-functional, a second path to the disk for the I/O request is identified as an alternative to the first path. The second path includes a second initiator endpoint and a second target endpoint and is identified by selecting a path from among those paths that have at least one endpoint that is different from the two endpoints of the first path.