Failure recovery method
The reliability is improved at a low cost even in a virtualized server environment. The number of spare servers is reduced for improving the reliability and for saving a licensing fee for software on the spare servers. A server system comprises a plurality of physical servers on which a plurality of virtual servers run, a single standby server, a module for detecting an active virtual server, and a module for switching the correspondence of boot disks of virtualization modules for controlling virtual servers to the physical servers. When a physical server fails, the boot disk of the associated virtualization module is connected to a spare server to automatically activate on the spare server those virtual servers which have been active upon occurrence of the failure.
1. A failure recovery method in a server system, the server system including a plurality of physical servers, wherein at least one of the plurality of physical servers executes a server virtualization module for providing a plurality of virtual servers, said method comprising the steps of:
operating the server virtualization module on a standby physical server;
providing correspondence information indicating a correspondence between each physical server and the server virtualization module executed on each physical server, and generating information regarding an active virtual server and a stopped virtual server on each server virtualization module;
capturing a snapshot of a memory image of the active virtual server and storing the snapshot in an image storage disk;
acquiring the information regarding the active virtual server and the stopped virtual server from the server virtualization module to update the correspondence information;
updating the information regarding the active virtual server and the stopped virtual server from the server virtualization module on a periodic basis or upon receipt of events; and
when one of the physical server fails: determining the virtual server which has been active on the server virtualization module executed on the failed physical server by referring to the correspondence information; and activating a virtual server on the server virtualization module executed on the standby physical server for recovering the determined virtual server using the snapshot of the determined virtual server stored in the image storage disk.
2. The failure recovery method according to claim 1 , further comprising the step of:
recording a location at which the snapshot is saved into the correspondence information,
wherein the snapshot is read from the location to recover the determined virtual server.
3. The failure recovery method according to claim 1 ,
wherein when another physical server fails after a virtual server has been activated on the standby physical server, a virtual server is generated in an active server virtualization module on the standby physical server to recover the virtual server in the another physical server.
4. The failure recovery method according to claim 1 , further comprising the step of:
recovering a virtual network set for a network of the virtual server when the virtual server is recovered.
5. The failure recovery method according to claim 3 , wherein when a new virtual server cannot be generated, another physical server is assigned to recover a virtual server.
6. A server system, comprising:
a plurality of physical servers,
wherein at least one of the plurality of physical servers executes a server virtualization module for providing a plurality of virtual servers, and
wherein the server virtualization module is operated on a standby physical server of the plurality of physical servers; and
a management server,
wherein the management server comprises:
a server virtualization module management table that provides correspondence information indicating a correspondence between each physical server and the server virtualization module executed on each physical server, and generates information regarding an active virtual server and a stopped virtual server on each server virtualization module;
a capture module that captures a snapshot of a memory image of the active virtual server and stores the snapshot in an image storage disk;
a virtual server information acquisition module that acquires the information on the active virtual server and the stopped virtual server from the server virtualization module to update the correspondence information, and updates the information on the active virtual server and the stopped virtual server from the server virtualization module on a periodic basis or upon receipt of events; and
a failure recovery module that, when one of the physical server fails, determines the virtual server which has been active on the server virtualization module executed on the failed physical server by referring to the correspondence information, and activates a virtual server on the server virtualization module executed on the standby physical server for recovering the determined virtual server using the snapshot of the determined virtual server stored in the image storage disk.
7. The server system according to claim 6 ,
wherein a location at which the snapshot is saved into the correspondence information is recorded, and
wherein the snapshot is read from the location to recover the determined virtual server.
8. The server system according to claim 6 ,
wherein when another physical server fails after a virtual server has been activated on the standby physical server, a virtual server is generated in an active server virtualization module on the standby physical server to recover the virtual server in the another physical server.
9. The server system according to claim 6 , wherein a virtual network set for a network of the virtual server is recovered when the virtual server is recovered.
10. The server system according to claim 8 , wherein when a new virtual server cannot be generated, another physical server is assigned to recover a virtual server.