Storage System Write Same Data Buffer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage systems incur excessive bus traffic and data fetch latency when repeatedly writing the same data, as they fetch the data from host memory multiple times for each write operation, which can be inefficient and disrupt system performance.
Innovation Solution
Implementing a dedicated write same data buffer in the storage system to store the data once fetched from host memory, allowing subsequent write operations to access it without re-fetching from host memory, thereby reducing bus traffic and enabling delayed execution of write commands to prioritize other operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data is fetched from host memory for each write operation, then write operations can be performed, but bus traffic increases and data fetch latency occurs
Solution Approach 1:
The patent applies preliminary action by fetching data from host memory to the storage system buffer before the actual write operations begin. This allows the data to be staged and ready for multiple sequential writes without repeated fetch operations, thereby reducing data fetch latency and improving write throughput.
Solution Approach 2:
The patent uses the storage system's internal buffer as an intermediary between host memory and the memory devices. Instead of directly fetching data from host memory for each write operation, the buffer serves as a mediator that holds the data locally, enabling efficient repeated writes to different logical addresses without increasing bus traffic.
2Reliability
If data is re-fetched from host memory for each write operation, then data accuracy is maintained, but bus traffic increases
Solution Approach 1:
The patent fetches data from host memory to the storage system buffer in advance before write operations begin. This preliminary fetch ensures data accuracy is maintained while avoiding repeated bus transactions during subsequent write operations, thereby reducing bus traffic and energy consumption.
Solution Approach 2:
The patent creates a local copy of the data in the storage system's buffer memory. This copy is then used for multiple write operations to different logical addresses, eliminating the need for repeated reads from host memory and reducing bus traffic while maintaining data accuracy through the buffered copy.
Data Source
AI summary
A command to duplicate data on a storage system is received from a host system, and in response to receiving the command, data corresponding to the command is retrieved from host memory of the host system. A plurality of write operations are performed on one or more memory devices of the storage system to fulfill the command, wherein performing each one of the plurality of write operations comprises writing the data to the one or more memory devices of the storage system without a subsequent retrieval of the data from the host memory.