Storage Event Logging and Visualization for Performance Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data storage devices often experience performance issues such as low throughput and high input/output (I/O) latency, which negatively impact quality of service (QoS) metrics and overall system performance, primarily due to inefficiencies in the storage application layers and hardware/software storage stacks.
Innovation Solution
Implementing a compute device with trace logic that captures and visualizes storage events across multiple application layers and data storage devices, allowing for time-synchronized logging and analysis of metrics like throughput and latency, enabling corrective actions to address performance issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If storage events are captured and logged across multiple application layers and data storage devices, then measurement precision of performance issues is improved, but device complexity increases
Solution Approach 1:
The patent introduces a compute device as an intermediary component that captures storage events from multiple application layers and data storage devices. This mediator consolidates the logging function, allowing precise performance measurement without requiring each individual storage device or application layer to independently implement complex logging mechanisms, thus improving measurement precision while managing system complexity through centralized coordination.
Solution Approach 2:
The logging system is segmented into distinct functional components: event capture logic at the storage device level, event collection at the compute device level, and analysis/separation of performance metrics. This segmentation allows each component to focus on specific tasks, improving the precision of performance issue detection while making the overall system more manageable and less complex through modular architecture.
2Reliability
If time-synchronized logging is implemented across multiple storage devices and application layers, then reliability of performance analysis is improved, but device complexity increases
Solution Approach 1:
The compute device implements time-synchronized logging by collecting storage events from multiple sources and analyzing them in coordination. The system uses timestamps and event correlation to ensure reliable performance analysis across distributed components. This feedback mechanism allows the system to adjust and correlate events from different layers, improving reliability of performance analysis while managing synchronization complexity through centralized event correlation logic.
3Productivity
If comprehensive storage event capture is implemented, then productivity of performance optimization is improved, but loss of time for data collection increases
Solution Approach 1:
The system implements preliminary action by continuously capturing and buffering storage events in the compute device before performance analysis is requested. This allows event data to be pre-collected and organized, so when performance optimization is needed, the analysis can quickly access already-captured data rather than starting collection from scratch, thus improving productivity while minimizing additional time loss.
Data Source
AI summary
Technologies for logging and visualizing trace capture data in a data storage subsystem (e.g., storage application layers and data storage devices of a compute device) are disclosed herein. One or more storage events in the data storage subsystem are captured for a specified time period. Statistics are determined from the captured storage events. A visualization of the storage events and statistics for the specified time period is generated.