Network Storage Bypassing CPU Bottleneck via Segmented Message Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional network-attached storage devices are inefficient due to the CPU bottleneck when processing data messages, leading to high power consumption and communication bandwidth requirements, as they handle both control and data messages, with data messages being the majority.
Innovation Solution
Implementing separate processing paths for control and data messages, where control messages are routed through the processing device for administrative and management operations, while data messages are directly communicated between the storage device and the storage client without going through the processing device, reducing the workload on the CPU and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the CPU processes both control and data messages in conventional network-attached storage devices, then the device can handle all storage operations, but the CPU becomes a bottleneck leading to high power consumption and limited scalability
Solution Approach 1:
The patent segments message processing into two separate paths: control messages are processed by the external processor through the host interface, while data messages are processed directly by the storage device through the network interface without involving the external processor. This segmentation eliminates the CPU bottleneck and reduces power consumption by offloading data processing from the external processor to the storage device itself.
2Ease of operation
If the external processor handles all messages including data messages, then centralized control is maintained, but communication bandwidth requirements increase and processing power needs are not reduced
Solution Approach 1:
The patent extracts data message processing from the external processor and assigns it directly to the storage device. The external processor only handles control messages through the host interface, maintaining centralized control for administrative operations while removing the data processing burden that would otherwise require high communication bandwidth and processing power.
3Device complexity
If data messages go through the processing device, then centralized processing is simplified, but security is reduced as data is exposed to the processing device
Solution Approach 1:
The patent segments the processing architecture into separate paths where data messages bypass the external processor entirely. The storage device processes data messages directly through the network interface, isolating data from the processing device and enhancing security while maintaining architectural simplicity through clear separation of control and data paths.
Data Source
AI summary
A standalone storage product having: a first bus connector for connecting to an external processor; a second bus connector for connecting to an external network interface; a storage device accessible over the network interface; and a processing device configured to communicate, via the second bus connector, with the network interface to obtain storage access messages represented by incoming packets received at the network interface from a computer network. The processing device can: identify, from the storage access messages, first messages and second messages; provide, the first messages via the first bus connector, to the processor; and provide, the second messages, to the storage device without the second messages going through the processor. The storage device is configured to: receive, via the first bus connector, third messages from the processor; and execute commands in the second messages and the third messages to implement a network storage service.