JPH09212468A - Compound mode multiprocessing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the information of trouble from a peripheral device which does not relate to operation by making a computer, allocated to the operation, monitor the processing capabilities of other computers and allocating an unallocated computer to the operation after preparing environment needed for the operation if one computer decreases below the specific processing capability. SOLUTION: When SSUI receives a request to report cluster information, a cluster management part 4 in the SSUI informs a CPU 6 in each cluster assigned to the operation of the cluster information. Then the processing capabilities of other CPUs 6 assigned to the same operation are monitored by a monitor program 6 and if one CPU decreases below the specific processing capability, an unassigned CPU 6 is assigned to the operation after the environment needed to executed the operation is prepared by actuating an OS 10. Consequently, the information of trouble from a peripheral device which does not relate to the operation can be suppressed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to dynamic allocation of computers to a plurality of jobs in a complex computer system in which a plurality of computers that execute a business program under the control of an operating system are connected.

[0002]

2. Description of the Related Art A conventional complex computer system or multiprocessor system monitors the load of a processor constituting a function when a plurality of processors perform a plurality of processes, as disclosed in, for example, Japanese Patent Laid-Open No. 4-235662. When overload is detected, a technique is disclosed in which a low-load processor is used as a processor having an overload function to efficiently perform load distribution.

[0003]

However, in order to execute one task, in addition to the CPU, the terminal device or the printing device is a hardware resource such as an external storage device, or software such as a file stored in the external storage device. Resources are needed. By the way, in order to manage these resources, it is necessary to arrange the management information in the memory, and in order to execute a plurality of functions with one processor, more memory is consumed. The memory that can be used for is reduced. In addition, as the number of managed peripheral devices increases, the frequency of failure notification will increase accordingly. Moreover, some of the notified failures are notified from the peripheral device which is not related to the work being executed. As a cause of these, there is a problem that the functional efficiency during the process is lowered.

[0004]

When a computer executes a certain task, the management information of resources not related to the task is not allocated in the memory, and a failure from a peripheral device not related to the task is prevented. It is necessary to suppress notifications.

[0005]

SUMMARY OF THE INVENTION The present invention solves the above problem by reconstructing only the environment required for a business when a computer is allocated to a business in a complex computer system.

[0006]

FIG. 1 is a block diagram of a system according to the present invention. 1 is an SSU, and multiple clusters are connected,
A shared storage area 2 shared by the connected clusters,
It consists of communication functions between clusters. In addition to SSU1,
It has a cluster management unit 3 that manages which section a cluster belongs to. The section is SSU1.
It is a unit for group management of clusters connected to.

Further, the shared storage area 2 has an initialization information management section 4 for managing the initialization information corresponding to each section. 5 is a cluster connected to SSU1, and C
It is composed of a PU 6 and a memory 7. Further, an external storage device 8 and a console 9 are connected to the cluster. The external storage device 8 has an operating system 10 (hereinafter referred to as an OS).
10), a business program 11 and a monitoring program 12 are stored. The business program 11 is executed under the control of the OS 10.

The process flow will be described with reference to FIG. When starting a job, the console 9 is used to specify a section number to start a cluster (step S2).
1). Here, the section number is represented by a positive number that corresponds to the job on a one-to-one basis. The activated cluster activates the OS 10 using the designated section number as the activation parameter (step S22). When booting the OS 10,
SS the initialization information corresponding to the specified section number
It is obtained from the initialization information management unit 3 of the common storage area of U1, and the hardware and software resources are initialized based on this. By this initialization, the management information and the like of the peripheral device are arranged in the memory 7 (steps S23 and S24). The configuration of the initialization information management unit 3 is as shown in FIG.

When the initialization is completed, the section number requested at the time of start is designated to the SSU 1 to request the section joining (step S25). In the SSU 1, when the section joining request is received, the cluster management unit 3 manages the fact. Next, the SSU 1 is requested to notify the section information (step S26). In the SSU 1, when requested to notify the cluster information, the cluster management unit 3 notifies the cluster information having the content shown in FIG. Specifically, the cluster number of the cluster that requested the cluster information notification,
It consists of the number of clusters connected to SSU1 and the section number to which each cluster belongs. A cluster that does not belong to any section is treated as an offline cluster, and the value in the cluster information is 0.

From the notified cluster information, it is judged whether or not it is the first cluster to which oneself has joined, depending on whether or not there is another cluster belonging to the same section as itself in the section, and if so, Will be the master cluster in that section. Otherwise, it becomes a slave cluster (step S2).
7). At this point it becomes the master cluster. The cluster that became the master cluster is the section monitoring program 12
Is activated (step S28). With the above, business processing is ready.

The flow of the section monitoring program will be described with reference to FIG. The section monitoring program 12 periodically obtains the CPU usage rate of clusters belonging to the same section and its average value (steps S41 and S4).
2) Based on that, cluster addition processing and cluster return processing are performed. If the average value of the CPU usage rate is 75% or more, cluster addition processing is performed (steps S43 and S4).
4) If the average value is less than 25%, cluster return processing is performed (steps S45 and S46).

The procedure of the cluster addition process will be described with reference to FIG. First, the SSU 1 is requested to notify the cluster information (step S51). When the SSU 1 receives the cluster information notification request, the cluster management unit 3 in the SSU 1
Notifies cluster information. The cluster number of the offline cluster and the section number of the section to which it belongs are obtained from the notified cluster information, and an IPL request notification request is made to the SSU 1 using these. If there is no offline cluster, no cluster can be added (steps S52 and S53).

Upon receiving the IPL request notification request, the SSU 1 makes an IPL request using the section number notified to the designated cluster as a parameter. The cluster requested for the IPL advances the processing according to the flow shown in FIG. The cluster started in step S21 starts the OS 10 using the specified section number as a start parameter (step S22). When the OS 10 is started, the initialization information corresponding to the specified section number is obtained from the initialization information management unit 3 in the common storage area of the SSU 1, and the hardware resources and the software resources are initialized based on this. By this initialization, the management information of the peripheral device and the like are arranged in the memory (steps S23 and S24).

When the initialization is completed, the SSU 1 is requested to join the section by designating the section number designated at the time of activation (step S25). In the SSU 1, when the section joining request is received, the cluster management unit 3 manages the fact. Next, the SSU 1 is requested to notify the section information (step S26). In the SSU 1, when requested to notify the cluster information, the cluster management unit 3 notifies the cluster information having the content shown in FIG.

From the notified cluster information, it is judged whether or not it is the first cluster to which the user has joined, depending on whether or not there is another cluster belonging to the same section as itself in the section, and if so, Will be the master cluster in that section. Otherwise, it becomes a slave cluster (step S2).
7). When started by another cluster, it always becomes a slave cluster, and step S28 is not executed. With the above processing, the job can be processed.

Next, the cluster return processing will be described with reference to FIG. The section management program 12 starts with the CP
Identify the cluster with the lowest U utilization. If it is not the master cluster, the cluster is set as a return target (steps S61 and S62). If the cluster identified in step S61 is the master cluster, it is confirmed whether or not there is a slave cluster in the section. If there is, the cluster with the second lowest CPU usage rate is identified.
The item is to be returned (steps S63 and S64). If there is a cluster to be returned, the cluster is notified of the business process transfer request, and if there is no cluster, nothing is done (step S65).

The cluster notified of the business process transfer request transfers the process being executed to another cluster in the same section and notifies the SSU 1 of an offline request.
The SSU 1 puts the cluster notified of the offline request into the offline state.

[0018]

When a job is changed, the environment is automatically set according to the job. Therefore, the environment to be managed can be the minimum required environment, and the management information and peripheral devices arranged in the memory can be used. Since the number of pieces of failure information to be notified is reduced, the overall processing efficiency is improved.

[Brief description of drawings]

FIG. 1 is a system configuration diagram of the present invention.

[Fig. 2] Flow chart of cluster startup

FIG. 3 is a diagram showing cluster information to be notified.

FIG. 4 is a diagram showing a flow of a section monitoring program.

FIG. 5 is a diagram showing cluster addition processing.

FIG. 6 is a diagram showing cluster return processing.

FIG. 7 is a diagram showing a configuration of an initialization information management unit.

[Explanation of symbols]

 1 SSU 2 Shared storage area 3 Initialization information management unit 4 Cluster management unit 5 Cluster 6 CPU 7 Memory 8 External storage device 9 Console 10 Operating system (OS) 11 Monitoring program 12 Business program

Claims (3)

[Claims] 1. A composite computer system in which a plurality of computers that execute business programs under the control of an operating system are connected, and each computer is assigned to any one of the businesses, or is assigned to any of the businesses. The computer assigned to one of the tasks executes the task process and monitors the processing capability of itself or another computer assigned to the same task, and the specified processing capability A computer system that allocates unallocated computers to the business after preparing the environment necessary to execute the business if 2. A complex computer system in which a plurality of computers that execute business programs under the control of an operating system are connected, and each computer is assigned to any one of the businesses, or is assigned to any of the businesses. The computer assigned to one of the tasks executes the task process and monitors the processing capability of itself or another computer assigned to the same task, and the specified processing capability If it is less than the specified value, the computer that has not been assigned is requested to start using the identification information of the environment required to execute the task as a parameter. A complex computer system that sets the environment based on the passed identification information of the environment. 3. A complex computer system in which a plurality of computers that execute a business program under the control of an operating system are connected, and the processing capacity of a computer assigned to a certain business is monitored and the processing capacity is predetermined. A compound computer system that has a monitoring unit that requests a computer that is not assigned to any job to start up by specifying the same environment as the job with insufficient capacity as a parameter when the processing capacity falls below.