The method of the automatic monitoring of system's unattended duty and device thereof
Technical field
The present invention relates to remote monitoring, cluster, system reducing technology, the method for the automatic monitoring of particularly a kind of system unattended duty.
Background technology
The automatic monitoring system is of many uses, also is the focus of studying at present.Particularly running up, needing the server end of process mass data simultaneously, the automatic monitoring technology of system seems particularly important.Existing remote control technology can be realized unattended operation, but system's handling failure automatically.The normal operation of maintenance system when a Clustering resolution system breaks down, but handling failure automatically.How realizing the automatic monitoring of unattended operation system, with the fastest speed and the highest timely treatment system fault of efficient, realize real unattended operation, reduce the loss that the system failure causes, is the problem that unattended operation system automatic monitoring Technology Need solves.
Common system's automatic monitoring technology comprises at present: 1, remote control technology; 2, Clustering; 3, manual system reduction technique.Brief account is following:
Prior art one: remote control technology
Principle: remote control technology mainly is made up of on-site supervision module, communication system and Surveillance center; The on-site supervision module is responsible for the collection of the information of accomplishing and the control command that send at the response monitoring center; Communication system is responsible for transmission of monitoring data and order; Surveillance center is responsible for collecting the monitor message that each monitoring module is uploaded, and sends various operational orders to monitoring module.
Shortcoming: only solved remote system unattended operation problem, can not repair automatically when breaking down, needed artificial repairing.
Prior art two: Clustering
Principle: cluster is a kind of parallel processing system (PPS), is made up of the independently computer that much links together, as the computational resource collaborative work of an integral body; Group system is meant that generally the two or more computer nodes that physically disperse link together through LAN, single system of picture for user and application program.
Shortcoming: only solved the problem of keeping system's operation when breaking down, and fault restoration still needs artificial treatment.
Prior art three: system reducing technology
Principle: the system reducing technology is exactly the original configuration of a system of backup in hard disk, when system breaks down, and the original configuration of recovery system.
Shortcoming: when system breaks down, be under people's operation, to reduce, rather than real automatic reduction.
In sum, prior art can not solve unattended system failure reparation, can not in time handle when system breaks down, and the cost of troubleshooting is high.
Summary of the invention
Instance of the present invention provides the method and apparatus of the automatic monitoring of a kind of system unattended duty, can not automatic monitoring in order to solve prior art, the defective of repairing automatically, and the high problem of system failure rehabilitation cost.
The method of the automatic monitoring of a kind of system unattended duty comprises:
Set up cluster virtual machine, through the mutual monitoring between the node, the fault point that discovery in real time can not normal access;
Trial is carried out the software fault reparation with the backup of virtual machine, the mode of system reducing;
Through the service of other virtual machine of timer access, confirm whether this virtual machine moves normally, in case can't visit, just make corresponding processing automatically.
The device of a kind of unattended operation system automatic monitoring comprises:
System has been divided into two-layer, is respectively the operation layer of cluster virtual machine composition and the supporting layer that virtual machine carrier (real equipment) cluster is formed;
The business one-tenth that virtual machine is trooped and formed is used for monitoring mutually between the node, and when certain node occurred repairing fault, this monitoring relation needed reorganization automatically;
The supporting layer that virtual machine carrier (real equipment) cluster is formed, the reduction request of sink virtual machine system.
Instance of the present invention is divided into the operation layer of cluster virtual machine composition and the supporting layer that virtual machine carrier (real equipment) cluster is formed to system; Monitor each other through certain logical relation between the virtual machine operation layer node, when certain node appearance can not be repaired fault, the monitoring relation reorganized automatically, and maintenance system normally moves, and real-time handling failure point; Except that hardware fault, software fault all can have been realized real unattended operation by system's automatic monitoring, reparation reduction automatically, has improved troubleshooting efficient, has reduced the troubleshooting cost.
Description of drawings
The unattended operation system braking monitoring that Fig. 1 provides for the embodiment of the invention, the method sketch map of handling automatically;
The apparatus structure sketch map of the unattended operation system automatic monitoring that Fig. 2 provides for the embodiment of the invention;
Fig. 3 is the environment sketch map of embodiment of the invention unattended operation system automatic monitoring;
Fig. 4 is the method flow diagram that embodiment of the invention unattended operation system virtual machine is handled automatically.