Memory Dump Flag System for Faster Fault Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing memory capacity in information processing systems leads to elongated memory dump times during faults, which can be costly to reduce through hardware improvements alone, and existing methods are inefficient in identifying and writing updated memory areas effectively.
Innovation Solution
The system employs a virtual storage system with a swap file and a flag storage unit to periodically write updated memory areas to a dump area, allowing for efficient pre-acquisition of data by clearing flags and using change bits or dirty bits to determine updated pages, thereby reducing the quantity of writes during faults.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If memory capacity is increased to handle larger data sets, then the system can store more information, but the memory dump time is elongated when faults occur
Solution Approach 1:
The physical memory is divided into multiple areas, each with its own flag. This segmentation allows the system to track and selectively dump only the updated areas rather than the entire memory space, reducing dump time while maintaining full memory capacity.
Solution Approach 2:
The system performs preliminary actions by setting flags in memory areas when they are updated during normal operation. This preliminary tracking enables the fault recovery process to quickly identify which areas need dumping without scanning the entire memory, thus reducing memory dump time while preserving large memory capacity.
2Speed
If hardware improvements are made to reduce memory dump time, then the dump speed increases, but the cost increases significantly
Solution Approach 1:
The patent introduces flags as intermediary markers in memory areas. These flags act as a low-cost software mechanism to identify updated memory regions, enabling fast selective dumping without requiring expensive hardware modifications. The flags provide a simple binary indication that guides the dump process efficiently.
3Reliability
If all memory areas are written to the dump area during faults, then complete memory state is captured, but the quantity of writes is excessive and time-consuming
Solution Approach 1:
The system extracts only the necessary information for fault analysis by using flags to identify and separate updated memory areas from non-updated ones. This extraction approach captures only the relevant memory state changes, ensuring complete fault analysis information while minimizing the quantity of writes to the dump area.
Solution Approach 2:
The flag mechanism provides feedback about which memory areas have been updated. This feedback system allows the dump process to make informed decisions about what data to write, ensuring that all necessary information for fault analysis is captured while avoiding unnecessary writes of unchanged data.
Data Source
AI summary
An apparatus including a physical memory partitioned into areas, a flag storage unit to store flags wherein a flag is set as indicating an area being updated when storage information stored in the area is updated, the area being associated with the flag, and a processor. The processor executes a first process of recording, when any fault does not occur, storage information stored in a first area to a recording device, the first area being associated with a first flag indicating the first area being updated, a process of saving, in a saving device, the storage information stored in the first area, and clearing the first flag so as to indicate the first area not being updated, and a second process of recording, to the recording device, storage information stored in a second area associated with a second flag indicating being updated when the fault occurs.