Network Adapter Metadata-Driven I/O Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data processing methods for network adapters require significant memory resources and increase processor overhead due to the need to migrate data between host and adapter memories, leading to increased program complexity and resource consumption.
Innovation Solution
The method involves processing I/O commands on the network adapter based on storage addresses and data lengths without requiring the actual data to be stored on the adapter, allowing data to be directly sent from the host server to remote servers without intermediate storage on the adapter, thereby reducing memory and processor load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is migrated from host memory to network adapter memory for processing, then data processing can be performed on the network adapter, but memory resource consumption and program complexity increase significantly
Solution Approach 1:
The patent extracts only the necessary metadata (address information and length information) from the data itself, leaving the actual data in host memory. This allows the network adapter to process storage services based on metadata without requiring large quantities of memory to store the actual data, thereby resolving the contradiction between data processing capability and memory resource consumption.
Solution Approach 2:
The patent uses metadata (address information and length information) as a copy or representation of the actual data characteristics. Instead of migrating physical data to the network adapter memory, the adapter works with this informational copy that contains all necessary processing instructions, significantly reducing memory requirements while maintaining processing functionality.
2Productivity
If data is migrated between host and network adapter memories, then data transfer can be performed, but processor overhead increases greatly
Solution Approach 1:
The patent extracts only the essential control information (address and length metadata) from the data stream, allowing the network adapter processor to make routing and transfer decisions without processing the actual data content. This dramatically reduces processor overhead while maintaining full data transfer capability.
Solution Approach 2:
The host system performs preliminary actions by preparing and sending metadata (address information and length information) along with data transfer requests. This allows the network adapter to be pre-instructed on how to handle the data without needing to analyze or process the data itself, reducing real-time processor overhead during data transfer operations.
3Productivity
If large quantities of memory are allocated on the network adapter for data storage, then data can be stored and processed, but program complexity increases
Solution Approach 1:
The patent extracts the essential storage control information (address and length metadata) from the data, allowing the network adapter to manage data storage operations using only this compact metadata. This eliminates the need for complex memory management programs and large memory allocations, while maintaining full data storage and processing capability through simplified metadata-driven operations.
4Productivity
If data is read from host memory to network adapter memory, then data can be processed on the adapter, but bus pressure increases greatly
Solution Approach 1:
The patent extracts only the minimal necessary information (address and length metadata) from the data, allowing the network adapter to process storage services based on this extracted information without requiring the actual data to be present in adapter memory. This eliminates the need for data migration over the bus, dramatically reducing bus pressure while maintaining data processing capability through metadata-driven operations.
Data Source
AI summary
A data processing method for a network adapter includes the network adapter that obtains a first input/output (I/O) command. The first I/O command instructs to store first data in at least one remote server to a local server, and the first I/O command includes address information and length information that are of the data and that are stored in the local server. The network adapter splits the first I/O command based on the address information and the length information to obtain a plurality of I/O read messages. The network adapter sends the I/O read messages to the at least one remote server.