Method and device for automatically detecting IO suspension abnormity of computing node

Abstract

The invention discloses a method and a device for automatically detecting IO suspension abnormity of a computing node, and relates to the technical field of cloud computing. The method comprises the following steps: collecting IO states of all virtual machines on the computing node in real time, wherein the IO states comprise a return state and a suspension state; counting the number of the IOs in the suspension state and the total number of the IOs at fixed time intervals, and determining the ratio of the number of the IOs in the suspension state to the total number of the IOs; and determining whether the computing node IO is in an abnormal state or not according to a size relationship between the ratio and a preset threshold value. According to the invention, IO abnormity of the computing node can be found in time, then effective processing measures are taken for the abnormal computing node in time, the response speed is increased, and the fault recovery speed is increased.

Description

technical field [0001] The invention relates to the technical field of cloud computing, in particular to a method and device for automatic detection of IO suspension abnormalities of computing nodes. Background technique [0002] This section is intended to provide a background or context to embodiments of the invention that are recited in the claims. The descriptions herein are not admitted to be prior art by inclusion in this section. [0003] In recent years, with the rapid development of cloud computing technology, its application has become more and more extensive. Generally, cloud platforms use distributed storage as storage resources to supply virtual machines. Since distributed storage can only detect faults within the cluster itself, the computing resources and When the storage connection is interrupted, the distributed storage cluster cannot judge the abnormal computing node by itself, causing the IO of the virtual machine on the abnormal computing node to be susp...

AI Technical Summary

Problems solved by technology

[0003] In recent years, with the rapid development of cloud computing technology, its application has become more and more extensive. Generally, cloud platforms use distributed storage as storage resources to supply virtual machines. Since distributed storage can only detect faults within the cluster itself, the computing resources and When the storage connection is interrupted, the distributed storage cluster cannot judge the abnormal computing node by itself, causing the IO of the virtual machine on the abnormal computing node to be suspended for a long time. The heartbeat of the virtual machine in the suspended IO state is normal, but it can no longer provide services normally.

[0004] At present, for scenarios such as computing node IO suspension, generally cloud vendors do not have a good processing mechanism. Basically, they rely on alarms to find out, manually operate and maintain down the computing node, and evacuate the above virtual machines to restore the environment. This processing method has Many problems such as slow response speed, difficult fault recovery and low efficiency

Images