Differentiated Alerting for Single Points of Failure in Distributed Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional single point of failure (SPOF) detection techniques in storage systems often lead to excessive false alarms, undermining storage system performance, particularly in distributed storage systems with multiple arrays.
Innovation Solution
Implementing a processing device that differentiates alert notifications based on the accessibility of logical storage devices across multiple storage arrays, generating distinct alerts for semi-SPOF and full-SPOF conditions, and applying tailored alert handling policies to reduce false alarms and improve system performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional SPOF detection techniques are used to monitor storage paths, then SPOF conditions can be detected, but excessive false alarms are generated which undermine storage system performance
Solution Approach 1:
The patent applies local quality by differentiating the alerting behavior based on the specific configuration context. Instead of using a uniform alerting approach for all SPOF detections, the system generates different types of alerts (full-SPOF vs. semi-SPOF) depending on whether the logical storage device is replicated across storage arrays in an active-active configuration. This localized differentiation reduces false alarms while maintaining reliable detection of actual failures.
2Measurement precision
If conventional SPOF detection techniques are used in distributed storage systems with multiple arrays, then path failures can be detected, but the number of false alarms increases excessively
Solution Approach 1:
The patent changes the alerting parameter based on the replication configuration state. When a logical storage device is replicated across multiple storage arrays in an active-active configuration, the system changes the alert type from full-SPOF to semi-SPOF. This parameter change in alert differentiation allows precise detection of actual SPOF conditions while filtering out false alarms that occur in replicated configurations.
3Device complexity
If uniform alert notifications are generated for all SPOF conditions, then simple alert handling is maintained, but false alarms cannot be distinguished from actual failures
Solution Approach 1:
The patent segments the alert notification system into two distinct alert types: full-SPOF alerts and semi-SPOF alerts. This segmentation is based on whether the logical storage device is replicated across storage arrays. By dividing the alerting mechanism into these two segments, the system maintains relative simplicity while significantly improving alert accuracy and enabling differentiation between false alarms and actual failures.
Data Source
AI summary
An apparatus comprises at least one processing device that includes a processor coupled to a memory. The processing device is configured to control delivery of input-output operations from a host device to at least first and second storage systems over selected ones of a plurality of paths through a network, to detect a single point of failure condition relating to a given one of the paths to a particular logical storage device in one of the first and second storage systems, and to determine whether or not the particular logical storage device is accessible in another one of the first and second storage systems. Different types of notifications are generated by the processing device depending on whether or not the particular logical storage device is accessible in the other one of the first and second storage systems.