Multipath server system fault diagnosis device, system and method

Abstract

The invention provides a multipath server system fault diagnosis device, system and method, the device comprises a Socket fault monitoring module and a storage module which are connected with each other, the Socket fault monitoring module is used for monitoring fault information of a CPU in each Socket in a multipath server system, and the monitored fault information is stored in the storage module. The Socket granularity fault diagnosis method can be suitable for a multipath server system to realize Socket granularity fault diagnosis, and has the advantages of simple structure, low complexity and cost, good expansibility and the like.

Description

technical field [0001] The invention relates to the technical field of multi-channel server systems, and in particular, to a fault diagnosis device, system and method for a multi-channel server system. Background technique [0002] The internal structure of a multi-channel server system is relatively complex. The system contains multiple CPU chips. Each CPU chip is interconnected through an interconnection channel (FIT) for data interaction. The internal resources of each CPU are shared. If cross-channel access occurs , the data needs to be interacted with through FIT. Therefore, in the multi-channel server system, it is very important to realize the internal fault diagnosis to ensure the stable and reliable operation of the system. [0003] Most of the fault diagnosis methods in the prior art are aimed at a single-channel system, and are usually implemented by adopting a software diagnosis method, that is, a software program is used to detect whether there is a fault in th...

AI Technical Summary

Problems solved by technology

[0004] 1. The diagnostic information acquisition process requires the support of software, which leads to the excessive dependence of the diagnostic process on the normal state of the software system. Once the software program is abnormal, the fault diagnosis of the entire system cannot be realized

[0005] 2. Due to the internal resource sharing of each CPU in the multi-channel server, even if the fault diagnosis based on the software diagnosis method detects a fault, it is difficult to locate the socket number where the fault occurred. At the same time, it is limited by the physical channel for information reporting. It is also impossible to report the fault information independently, that is, the software diagnosis method is difficult to refine the fault diagnosis to the socket granularity

[0006] In the prior art, there are few studies on the fault diagnosis of multi-channel server systems, and the fault diagnosis is usually realized by analyzing the fault log information, and the fault discovery of the CPU small system is not paid attention to, and different sockets cannot be distinguished from each other. diagnosis

In the prior art, it is necessary to rely on the design of auxiliary external circuits to realize the fault discovery of small CPU systems, such as figure 1 As shown, the external circuit is used to capture the fault signal and analyze the fault information, but this will not only increase the difficulty of board-level design, but also make the system difficult to generalize

Images