Anomaly-Aware Log Retrieval for Disk Array Enclosures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current disk array enclosures (DAEs) have limited ring buffer sizes for log storage, leading to rapid overwriting of logs, making it difficult for administrators to troubleshoot issues as critical log data is often lost before it can be retrieved and analyzed.
Innovation Solution
Anomaly aware log retrieval systems that include a log retrieval module capable of detecting anomalies in storage systems, automatically triggering log collection and classification, and storing logs in a host server's local storage to prevent overwriting, ensuring that logs associated with specific anomalies are retained.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If logs are stored in a small buffer at the DAE, then storage space is conserved, but logs are overwritten quickly making them unavailable for analysis
Solution Approach 1:
The patent extracts logs from the DAE's internal buffer and transfers them to external storage locations including the host server's file system, remote storage systems, and cloud storage. This separation allows the small internal buffer to continue functioning while logs are preserved externally for long-term analysis and troubleshooting.
Solution Approach 2:
The system proactively collects and stores logs before they are overwritten by implementing continuous monitoring, anomaly detection, and automated log retrieval mechanisms. When anomalies are detected or periodically, logs are immediately captured and transferred to external storage, ensuring log availability before the overwrite occurs.
2Loss of information
If administrators manually retrieve logs frequently, then log availability increases, but time and operational effort increase
Solution Approach 1:
The system implements automated log retrieval and storage mechanisms that operate without administrator intervention. The log management module continuously monitors the DAE, automatically detects anomalies, retrieves logs, and stores them in external locations, eliminating the need for manual log collection and reducing administrative time and effort.
Solution Approach 2:
The system continuously monitors DAE operations and log buffer status, providing feedback to trigger automated log retrieval when anomalies are detected or when the buffer is nearing capacity. This feedback mechanism ensures logs are captured at optimal moments without requiring constant administrator attention.
3Duration of action of stationary object
If the buffer size is increased at the DAE, then logs are retained longer, but storage capacity is consumed and cost increases
Solution Approach 1:
Instead of increasing storage capacity in the traditional horizontal dimension (larger buffer), the patent transitions to a hierarchical storage architecture utilizing multiple dimensions: fast local buffer at DAE, host server file system, remote storage systems, and cloud storage. This multi-dimensional approach extends log retention time without requiring proportional increases in any single storage capacity.
Data Source
AI summary
An anomaly with a disk array enclosure (DAE) of a set of DAEs connected to a host server is detected. Each DAE is chained to another DAE by first and second chains and includes a serial attached SCSI (SAS) expander and a peer SAS expander. The SAS expander in a DAE forms part of the first chain. The peer SAS expander in the DAE forms part of the second chain. Upon detection, DAE log collection is triggered to obtain logs for storage in a repository, separate from the DAEs. The logs are tagged with an anomaly class of a set of anomaly classes. For each anomaly class, a number of logs stored in the repository is limited to a threshold number.