Server memory failure alarm method

Abstract

The invention provides a server memory failure alarm method. A software program is used for recognizing memory error messages at a startup stage of a server system, the messages are transmitted to a management chip embedded into a main board and are judged in a classified manner, and alarm is given according to levels. The system comprises a failure information recognition unit, an error message database and an alarm unit, wherein the failure information recognition unit is used for acquiring the error messages sent out by a memory component in the system to provide evidence for judging memory failure, the error message database is used for collecting and transmitting the memory error messages, and the alarm unit is used for selecting different failure alarm modes according to different error messages. By implementing the memory failure alarm method in the server system, reliability of the system can be enhanced to a large extent, maintenance is facilitated, and the total image of the system is improved.

Description

technical background [0001] In the current server system, the existing technology only triggers the hardware fault circuit signal through the memory component, and performs fault indication through the onboard LED. This type of design has the following shortcomings: [0002] 1. The fault information cannot be recorded. Once the system is powered off, the fault information found during this startup will be cleared; [0003] 2. The types of faults that the system can identify are limited: only simple error messages that can be detected by the memory components themselves are supported, such as the memory temperature is too high and the number of Error IO records is too high. However, for error messages that cannot be detected or reported by memory components such as memory chip failures and memory installation errors, the server system cannot generate alarm messages; [0004] 3. It is impossible to report to the police according to the severity of the fault. Contents of the i...

AI Technical Summary

Problems solved by technology

[0002] 1. The fault information cannot be recorded. Once the system is powered off, the fault information found during this startup will be cleared;

[0003] 2. The types of faults that the system can identify are limited: only simple error messages that can be detected by the memory components themselves are supported, such as the memory temperature is too high, and the number of Error IO records is too high

However, for error messages that cannot be detected or reported by memory components such as memory chip failures and memory installation errors, the server system cannot generate alarm messages;

[0004] 3. It is impossible to alarm according to the severity of the fault

Images