Out-of-band I/O Module Recovery via Network Adapter
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage systems lack a means to independently control and recover an unresponsive I/O module, leading to potential system downtime due to the lack of an out-of-band control path.
Innovation Solution
Implementing an out-of-band control path that allows a storage server to send control commands, such as reset or power cycle commands, to an I/O module through a network adapter, independent of the in-band data path, to recover a non-responsive I/O module and obtain its status data for analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If only an in-band data path is used to communicate with the I/O module, then the system structure is simple, but the system cannot independently control or recover a non-responsive I/O module
Solution Approach 1:
The patent segments the communication path into two independent channels: in-band data path for data transfer and out-of-band control path for management and recovery. This segmentation allows the control function to operate independently from the data path, enabling recovery of non-responsive I/O modules without affecting data operations.
Solution Approach 2:
The patent introduces a management processor as an intermediary component that receives control commands through the out-of-band control path and executes them on the I/O module. This intermediary enables indirect control of the I/O module when direct control through the data path is unavailable due to module unresponsiveness.
2Productivity
If an out-of-band control path is implemented, then automatic recovery of non-responsive I/O modules is enabled, but the system complexity increases
Solution Approach 1:
The patent implements preliminary actions by establishing the out-of-band control path and management processor before failures occur. When an I/O module becomes non-responsive, the system can immediately execute recovery commands (reset, reinitialization) through the pre-established control path without needing to reconfigure or diagnose the failure mode first.
Solution Approach 2:
The patent implements feedback mechanisms where the management processor monitors the status of I/O modules through the out-of-band control path and automatically executes recovery actions when anomalies are detected. This closed-loop feedback enables automatic recovery without administrator intervention, improving system availability.
3Loss of time
If the storage server waits for I/O module response through the in-band path, then no additional control infrastructure is needed, but system downtime occurs during I/O module failures
Solution Approach 1:
The patent adds a new dimension to the communication architecture by introducing the out-of-band control path that operates independently from the in-band data path. This dimensional addition allows control operations to proceed parallel to data operations, eliminating the time loss that would occur if the system had to wait for I/O module responsiveness through the data path.
Data Source
AI summary
A storage system includes a storage server and a storage subsystem having an I/O module coupled to disks. The storage server includes a storage adapter coupled to the I/O module creating an in-band (data) path to access data stored on the disks. The storage server sends a data command through the in-band path to the I/O module for accessing the data stored on the disks. The storage server further includes a network adapter coupled to the I/O module creating an out-of-band control path to control the I/O module. The storage server detects through the in-band path that the I/O module is non-responsive to a data command and automatically changes the behavior of the I/O module (e.g., resets the I/O module) through the out-of-band control path to recover the non-responsive I/O module. The storage server also collects and outputs status data of the I/O module through the out-of-band control path.