Software-Based Write-Through Caching for Redundant Controllers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial process control and automation systems face challenges in maintaining redundant control operations without significant downtime, especially when primary controllers fail, due to limitations in hardware-based write-through caching modes, which are not supported by all processor architectures.
Innovation Solution
Implementing a software-based write-through caching method using redundant data types with tracked operators that include assignment and transfer functions to mirror data from primary memory to a backup memory, ensuring seamless transition in case of primary controller failure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If hardware-based write-through caching is used, then data transfer speed and redundancy reliability are improved, but device compatibility and ease of manufacture deteriorate because not all processor architectures support this mode
Solution Approach 1:
The patent replaces the hardware-based write-through caching mechanism with a software-based implementation. The software simulates write-through caching behavior by intercepting write operations and manually copying data from primary memory to backup memory, thereby achieving the same redundancy effect without requiring specific hardware support. This substitution resolves the contradiction by maintaining reliability while improving adaptability across different processor architectures.
Solution Approach 2:
The patent introduces a software layer as an intermediary between the processor and the memory system. This software intermediary implements the write-through caching logic, acting as a mediator that translates standard write operations into coordinated updates across primary and backup memory. This intermediary approach allows the system to achieve hardware-level reliability through software without being constrained by processor architecture limitations.
2Loss of time
If real-time data mirroring is implemented, then downtime is reduced, but data transfer overhead and system complexity increase
Solution Approach 1:
The patent implements selective data mirroring rather than complete real-time copying of all memory operations. The software monitors and identifies critical data changes that require immediate replication to backup memory, while allowing non-critical data to be updated asynchronously or batched. This partial action approach reduces the overall data transfer overhead while still maintaining sufficient redundancy to minimize downtime during failures.
Solution Approach 2:
The patent pre-allocates backup memory space and pre-configures the software-based write-through caching mechanism before failures occur. By having the redundancy infrastructure ready in advance, the system can immediately switch to backup memory when needed, minimizing downtime without requiring complex real-time data transfer protocols. The preliminary setup reduces the complexity of emergency data synchronization.
Data Source
AI summary
An industrial process control system comprises a primary controller coupled with a buffer and a primary memory. The primary controller is configured to create at least one redundant data type based on at least one base data type. The at least one redundant data type includes tracked operators that are different from corresponding operators of the at least one base data type. The tracked operators include an assignment function that is a same assignment function performed by the corresponding operators of the at least one base data type, and a transfer function. The primary controller is further configured to perform the assignment function on a piece of data to assign the piece of data to an address in the primary memory, perform the transfer function to transfer the piece of data to a buffer, and transfer contents of the buffer to a backup memory.