Fault detection method, device and equipment for memory in server

Abstract

The invention provides a fault detection method, device and equipment for a memory in a server, and the method comprises the steps: mapping a memory address in the server into a memory space, carrying out the division processing of the memory space, and carrying out the fault test of each divided memory block, selecting a target memory address read-write mode from a plurality of different memory address read-write modes associated with the memory block, and selecting a target pressure measurement algorithm from a plurality of associated different pressure measurement algorithms; carrying out pressure measurement detection on the memory block according to a target memory address read-write mode and a target pressure measurement algorithm, and if the memory block is detected to have a fault, recording the number of memory faults of the memory block; and when the accumulation of the fault records corresponding to the memory blocks reaches the configured memory fault report threshold value, ending the pressure measurement detection of the memory blocks, and reporting the fault information of the faulted memory blocks. Therefore, according to the embodiment, the detection efficiency can be improved, and meanwhile, the memory fault detection rate can be increased.

Description

technical field [0001] The present application relates to a fault detection technology, and in particular, to a fault detection method, device and device for a memory in a server. Background technique [0002] For servers, memory, motherboard, and CPU are the three major components that make up a server. Among these three components, server failures caused by memory account for the highest proportion of the three components. The failure caused by the memory will cause the server to go down with high probability, causing the interruption of the running business and causing huge economic losses. Especially for servers sold abroad, once a failure occurs, maintenance is extremely difficult due to the long distance, which brings great pressure to the manufacturer's after-sales service. Based on this, improving the detection rate of faulty memory on the production side is of great significance to the quality of the server. [0003] At present, all virtual addresses allocated to ...

AI Technical Summary

Problems solved by technology

However, the existing technology generally improves the detection rate of the memory by increasing the pressure measurement algorithm of the test memory. There are many types of pressure measurement algorithms for the current test memory, but the basic technology is mature, and it is difficult to make a breakthrough.

Therefore, in order to improve the detection rate of production server memory, we can only continuously increase the memory pressure test time, which leads to low detection efficiency of memory detection and greatly increases production costs.

Images