Ceph-based virtual machine high availability implementation method and system

Abstract

The invention, which relates to the technical field of cloud computing, discloses a Ceph-based virtual machine high availability implementation method and system. The method comprises: a control nodeserver sends virtual machine high availability configuration information to all computing node servers in real time; after each computing node server obtains an exclusive lock of a mirror image in real time, a first heartbeat timestamp of a virtual machine using the mirror image is recorded; if the exclusive lock of the mirror image can not be obtained and the first heartbeat of the virtual machine using the mirror image exceeds the time, the virtual machine is closed; the control node server breaks the exclusive lock of the mirror image that has been acquired regularly after or while the virtual machine is closed and a second heartbeat timestamp of the virtual machine using the mirror image is recorded; and when the exclusive lock of the mirror image is not obtained and the second heartbeat of the virtual machine using the mirror image exceeds the time, the virtual machine using the mirror image is dispatched to another computing node server for and is started again. Therefore, virtual machine high availability is realized by detecting the heartbeat of the virtual machine.

Description

technical field [0001] The invention relates to the technical field of cloud computing, in particular to a Ceph-based virtual machine high availability implementation method and system. Background technique [0002] In cloud computing data centers, virtual machines usually run on physical servers. When the physical server is powered off or has other failures, the virtual machines running on it will also be shut down, resulting in unexpected interruption of user services running in the virtual machines. To ensure automatic recovery of user services after unexpected interruptions, it is necessary to monitor the status of virtual machines in real time. When a physical server fails, transfer the virtual machines on it to other physical servers for restart. This is called virtual machine high available. [0003] In traditional data centers, the high availability of virtual machines requires the use of shared storage to store the disk files of virtual machines, so that all physic...

AI Technical Summary

Problems solved by technology

Creating lock space files and resource files on distributed storage will also be divided into many objects and stored on n physical server disks, and the objects of lock space files, resource files and virtual machine disk files are not necessarily mapped to On the same physical server disk, when the copy of the heartbeat lock is accessible, it does not mean that the copy of the virtual machine disk file is accessible, so the shared storage solution cannot be used to achieve high availability of the virtual machine by detecting lock space files and resource files

Images