Storage Drive Telemetry Tap for Selective Streaming Alerts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage device telemetry data collection is either too coarse or too clunky to be useful for online data collection, leading to difficulties in debugging and prediction of drive failures, or results in excessive data that is difficult to process into actionable information.
Innovation Solution
A streaming data interface, or 'telemetry tap', is used in conjunction with a host-defined telemetry program to regulate the type and amount of telemetry data sent to the host device, allowing the controller to execute host-generated programs that define which telemetry operations should occur and be forwarded via the streaming telemetry mechanism.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If SMART log format is used for telemetry data collection, then the operation is lightweight and preferred by host and drive, but the data is often inaccurate or incomplete
Solution Approach 1:
The patent combines multiple telemetry data sources (SMART logs, vendor-specific internal logs, and streaming data interfaces) into a unified telemetry data collection system. The controller aggregates data from these different sources to provide both lightweight operation and accurate, complete telemetry information, resolving the contradiction between operation weight and data accuracy.
2Loss of information
If vendor specific internal log pulling is used, then complete telemetry data can be obtained, but the operation is heavy and impacts device and host performance
Solution Approach 1:
The patent extracts only the necessary telemetry data from the vendor-specific internal logs based on host requests, rather than pulling all data. The controller selectively retrieves specific telemetry parameters needed for monitoring and debugging, reducing the operational burden while maintaining data completeness for the requested parameters.
Solution Approach 2:
The patent implements a dynamic telemetry data collection mechanism where the host can programmatically request specific telemetry data based on current needs. The controller adapts its data collection behavior according to host requests, suspending operations only when absolutely necessary and resuming when not impacting performance, thus dynamically balancing data completeness with system performance.
3Loss of information
If all telemetry data is collected and provided to host, then complete information is available, but the customer has no means to reduce the collected data into actionable information
Solution Approach 1:
The patent introduces an intermediary processing layer in the controller that filters, aggregates, and processes telemetry data before presenting it to the host. This intermediary layer reduces the raw telemetry data into actionable information while maintaining completeness of critical data, thereby reducing the complexity of data processing required by the customer.
Solution Approach 2:
The patent performs preliminary processing of telemetry data in the controller, including filtering, aggregation, and formatting, before transmitting data to the host. This preliminary action reduces the data volume and complexity that the customer must process, while ensuring that all necessary information is already prepared and organized for actionable use.
4Loss of information
If vendor specific internal log pulling mechanism is used, then telemetry data can be obtained, but other operations must be suspended and user workloads are impacted
Solution Approach 1:
The patent implements a dynamic telemetry data collection mechanism that adapts to system conditions. The controller monitors system state and selectively suspends or continues telemetry operations based on current workload and performance conditions. This dynamic approach ensures telemetry data accessibility while minimizing impact on user workloads by performing data collection only when it does not interfere with normal operations.
Data Source
AI summary
A streaming data interface or a ‘telemetry tap’ in conjunction with a host defined telemetry program is used to regulate the type and amount of telemetry data sent to the host device. The amount of telemetry data provided to the host is based on a request. The controller will receive and execute valid host generated programs which define which telemetry operations should occur and be forwarded to the host via the streaming telemetry mechanism. The controller will use the user/host programmable mechanisms that will collaborate with internal drive logging mechanisms. The controller will watch for the host-defined programmable mechanisms and send the requested amount of telemetry data to the host once the programmable mechanisms have executed.