JP2001067331A - Nonstop modifying method of program of multi-processor system and multi-processor system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To modify a program to be executed by a group of processors without fully stopping the group of processors. SOLUTION: In this multi-processor system, the group of processors (211, 212, 213, 214, 221, 222, 223, 224) is grouped into two crusters (a cruster 1, a cruster 2), a distributed processing of a job is normally performed irrespective of the crusters by all processors, when a program is modified, a mode is transited to the one to perform the distributed processing of the job only between processors of the same cruster first, a processing at the cruster 1 is suspended as continuing a processing at the cruster 2, programs of respective local memories LM of the processors (211, 212, 213, 214) of the cruster 1 are modified, next, the processing at the cruster 2 is suspended as continuing the processing at the cruster 1, programs of respective local memories LM of the processors (221, 222, 223, 224) of the cruster 2 are modified, the processing is restarted after modification and after that, the processors are restored to normal operation.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a non-stop program change method for a multiprocessor system and a multiprocessor system. More specifically, the present invention relates to a non-stop program change method for a multiprocessor system and a multiprocessor system that can change a program executed by a processor group without stopping the entire processor group.

[0002]

2. Description of the Related Art Non-stop maintenance in which a main processor and a sub processor are provided, a backup system is switched to take over the processing performed by the main processor by a sub processor, and then the main processor is stopped and maintained. The method is described, for example, in JP-A-56-52460.

[0003]

According to the conventional non-stop maintenance method described above, it is possible to maintain a system portion belonging to the main processor without stopping the entire process. However, in order for the processing performed by the master processor to be taken over by the slave processor, the programs executed by the master processor and the slave processor must be the same version. There is a problem that the above non-stop maintenance method cannot be applied. That is, in the conventional technique, there is a problem that when the program is upgraded, the entire processor group must be stopped. Therefore, an object of the present invention is to stop the entire processor group without stopping.
An object of the present invention is to provide a non-stop program change method for a multiprocessor system that can change a program executed by a processor group and a multiprocessor system.

[0004]

According to a first aspect of the present invention, a program is stored in each local memory of a plurality of processors, and each processor executes the program stored in its own local memory to execute a job. In a multiprocessor system for processing, the plurality of processors are grouped into N clusters from a cluster 1 to a cluster N (N is a natural number equal to or greater than 2 and equal to or less than the total number of processors). A cluster limited mode for performing job distribution processing only between processors belonging to the same cluster and a cluster non-limited mode for performing job distribution processing among all processors are provided. When jobs are distributed and the program is changed, the non-cluster From the processing of the job in (1) to the processing of the job in the cluster limited mode. When the migration is completed, one cluster is selected, and the processing of the job and the acceptance of a new job in the processor group of the non-selected cluster are continued. Stop accepting a new job in the processor group of the selected cluster, and when there are no more jobs to be processed in the processor group of the selected cluster, replace the programs stored in each local memory of the processor group of the selected cluster with a new program. After changing the program of all the processors of the selected cluster, the process of accepting a new job and the processing of the job in the processor group of the selected cluster is restarted, and this is repeated for all the clusters in order. Job processing in cluster non-restricted mode. Providing uninterrupted program changing the multiprocessor system back to the process, characterized in that return to normal operation.

In the method for changing a non-stop program of a multiprocessor system according to the first aspect, a group of processors is divided into two or more clusters. At first, shift to the mode of distributing jobs only between processors of the same cluster, select one cluster when the shift is completed, and suspend the processing in the selected cluster while continuing the processing in the non-selected cluster. To change the program of the processor of the selected cluster, restart the processing after the change, and repeat this for all clusters.
After the change of the program and the resumption of the processing are completed for all the clusters, the processing returns to the normal operation in which the jobs are distributed and processed by all the processors regardless of the cluster. Thus, the program can be changed without stopping the operation of the multiprocessor system entirely and without interfering with different versions of the program.

According to a second aspect, the present invention relates to a multiprocessor system in which a program is stored in each local memory of a plurality of processors, and each processor executes a program stored in its own local memory to process a job. , The plurality of processors are grouped into N clusters from cluster 1 to cluster N (N is a natural number not less than 2 and not more than the total number of processors),
A cluster limited mode in which the cluster group is divided into two sets, a first set and a second set, and job distribution processing is performed only between processors belonging to the same cluster as the cluster to which the own processor belongs; A non-cluster limited mode for performing a job distribution process between a processor group belonging to the same cluster as the cluster to which the own processor belongs and a processor group belonging to the same group as the set to which the own processor belongs. ,
At the time of the normal operation, all the clusters belong to one of the first group and the second group, and the jobs are distributed in the non-cluster limited mode. At the time of the program change operation, first, one cluster is selected and unselected. The job processing in the non-cluster-only mode in the selected cluster is shifted to the job processing in the cluster-only mode in the selected cluster while the distributed processing of the job in the non-cluster-limited mode in the cluster is continued. Stop accepting new jobs in the group, and when there are no more jobs to be processed by the processors in the selected cluster, change the programs stored in each local memory of the processors in the selected cluster to new programs, and select After completing the program change for all processors in the cluster, select the processors in the cluster. Accepts a new job and resumes job processing, changes the affiliation of the selected cluster from one of the first set or the second set to the other, and then processes the job in the cluster limited mode in the selected cluster. A method for changing a non-stop program of a multiprocessor system is provided, wherein the method returns to the job processing in the non-limited mode, and repeats this for all clusters to return to the normal operation.

In the method for changing a non-stop program of a multiprocessor system according to the second aspect, a group of processors is divided into two or more clusters. At first, shift to the mode in which jobs are distributed only between processors in the same cluster and processors in a predetermined cluster that can be distributed, select one cluster when migration is completed, and continue processing in non-selected clusters While the processing in the selected cluster is paused, the program of the processor in the selected cluster is changed, the processing is restarted after the change, this is repeated for all clusters, and the change of the program and the restart of processing are completed for all clusters And then distribute jobs to all processors regardless of the cluster. To return to normal operation. As a result, the program can be changed without stopping the operation of the multiprocessor system entirely, without interfering with programs of different versions, and while performing distributed processing among processors of different clusters.

In a third aspect, the present invention provides a multiprocessor system for implementing the non-stop program changing method according to the first aspect or the non-stop program changing method according to the second aspect.

In the multiprocessor system according to the third aspect, since the above-described non-stop program change method can be suitably implemented, it is not necessary to stop the entire operation when changing the program.

[0010]

Embodiments of the present invention will be described below with reference to the drawings. Note that the present invention is not limited by this.

First Embodiment FIG. 1 is a configuration diagram of a disk control device 200 according to a first embodiment of the present invention. This disk control device 200
Are processors 21 each having a local memory LM
1,212,213,214,221,222,223
And 224 and the control program storage drive 21
Service processor 21 having each of
5, 225 and a shared memory 230 having a cache memory 231 and a control memory 232.

That is, the disk controller 200
Are the eight processors 211, 212, 213, 21
4, 221, 222, 223 and 224.

The eight processors 211, 212, 2
In the local memories LM of 13, 214, 221, 222, 223 and 224, a data transfer control program and a maintenance support program are stored. Each of the processors 211, 212, 213, 214, 221, 22
2, 223 and 224, the data transfer control program is executed, and data transfer control processing is performed (that is, the disk control device 200 is a multiprocessor system). Further, the maintenance support program is executed, and non-stop maintenance processing is performed. Processors 211, 212, 221 and 222 in the processor group
Is the channel CP connected to the channel controller CC
U. Therefore, these are called channel-side processors. Further, the processors 213, 214, 223 and 224 include a drive CPU connected to a disk drive (hereinafter simply referred to as a drive) D0. Therefore, these are called drive-side processors. Further, the processor group is divided into two clusters. Processors 211, 212, 213 and 214 are "cluster 1"
222, 223 and 224 are “cluster 2”.
The service processors 215 and 225 are connected to the maintenance terminal 2
40. In addition, the service processor 21
5 belongs to the “cluster 1”, and the service processor 2
Reference numeral 25 belongs to the "cluster 2".

The control memory 232 stores cluster allocation information 500 as shown in FIG. In the processor number area of the cluster assignment information 500, the number of each processor is stored (in the present embodiment, 2).
11, 212, 213, 214, 221, 222, 22
3 and 224). In the belonging cluster number area, the cluster number of the cluster to which the corresponding processor belongs is stored (in this embodiment, the processor 21).
1, 212, 213 and 214 are cluster numbers "1"
And the processors 221, 222, 223 and 22
4 is the cluster number “2”). In the processor attribute area, information indicating the function of each processor in the distributed processing of the data transfer job is stored (in this embodiment, the processors 211, 212, 221 and 22).
2 is a “channel CPU”, and the processors 213 and 2
14, 223 and 224 are "drive CPUs").

The control memory 232 stores execution cluster instruction information 600 as shown in FIG. In the execution cluster instruction area of the execution cluster instruction information 600, “1 (cluster non-limited mode)” is stored when both the cluster 1 and the cluster 2 can be used for data transfer. If only one of them can be used for data transfer, "2 (cluster limited mode)" is stored.

The control memory 232 stores I / O job execution control information 8 as shown in FIG. 4 created for all drives connected to the disk control device 200.
00 is stored. This I / O job execution control information 8
00, the command type (read or write) of the access command and the drive number of the access target (I / O job execution control information 80
(Fixed every 0), and the cylinder number, track number, and record number of the access target. In addition, the processor 211, 212, 213, 214, 221, 222, 2
23 and 224 store information indicating whether or not the access command can be executed (“1” if executable, “0” if not executable). In the I / O request execution state area, information indicating the execution state of the access command for the drive (“0” if there is no execution request, drive CP
"1" is stored if the U is waiting for execution, "2" is stored if the CPU is waiting for execution, and "3" is stored if any processing is being performed.

Further, in the control memory 232, FIG.
The cluster operation mode instruction information 700 as shown in FIG. The cluster number (“cluster 1”, “cluster 2”) is stored in the cluster number area of the cluster operation mode instruction information 700. In the operation mode area, “1 (normal operation mode)” indicating a normal operation mode, “2 (pause mode)” indicating an operation mode in which an access command is not accepted, or cluster 1
And “3 (maintenance operation mode)” indicating that the cluster 2 and the cluster 2 are executing different versions of the data transfer control program.

Next, the data transfer control processing of the disk controller 200 will be described. When an access command is passed from a host computer (not shown) to the channel controller CC, the channel controller CC transmits the access command to the cluster 1 (21) of the disk controller 200.
0) or one of the channel-side processors 211, 212, 221, 222 of the cluster 2 (220). Here, it is assumed that the access command is a read request to the drive D0, and that the access command is passed to the channel-side processor 211 of the cluster 1 (210).

The channel-side processor (211) to which the access command has been passed receives the I / O corresponding to the drive D0.
Only when "0 (no execution request)" is set in the I / O request execution status area of the job execution control information 800 (FIG. 4), "3 (executing)" is displayed in the I / O request execution status area. Is stored in the access command request area, and "1 (drive CPU execution waiting)" is set in the I / O request execution state area. Next, the channel-side processor (211) refers to the execution cluster instruction information 600 (FIG. 3), and when the execution cluster instruction area is “1 (cluster non-limited mode)”, the I / O job execution control The drive-side processor numbers (213, 214, 223, 2) in the executable processor designation bitmap area of the information 800 (FIG. 4)
24. “1 (executable)” is set in the execution possibility field corresponding to (24). On the other hand, when the execution cluster instruction area is “2 (cluster limited mode)”, the local processor (here, the channel side) in the executable processor designation bitmap area of the I / O job execution control information 800 (FIG. 4). "1 (executable)" is set only in the execution possibility field corresponding to the drive-side processor number (213, 214) belonging to the same cluster (here, cluster 1) as the processor 211).

Drive side processors (213, 214,
223, 224) (here, the drive-side processor 223) refers to the I / O request execution status area of the I / O job execution control information 800 (FIG. 4) and sets “1”. (Drive CPU execution waiting) "is set, refer to the execution availability field corresponding to the own processor number (223) in the executable processor designation bitmap area of the I / O job execution control information 800 (FIG. 4). Only when "1 (executable)" is set therein, "3 (executing)" is set in the I / O request execution state area, and the information ( Data is read out according to the drive number, cylinder number, track number, record number), stored in the cache memory 232 of the shared memory 230, and Control information 800 "2 (channel CPU execution wait)" in the I / O request execution state area (Fig. 4)
Is set. Then, the execution cluster instruction information 600
Referring to FIG. 3, when the execution cluster instruction area is “1 (cluster non-limited mode)”, the executable processor designation bitmap area of the I / O job execution control information 800 (FIG. 4) Is set to "1 (executable)" in the executable / non-executable field corresponding to the channel-side processor number (111, 212, 221, 222). On the other hand, when the execution cluster instruction area is “2 (cluster limited mode)”, the I / O job execution control information 8
00 (FIG. 4) in the executable processor designation bitmap area.
"1 (executable)" is set only in the execution availability field corresponding to the channel-side processor numbers (221, 222) belonging to the same cluster (cluster 2) as in 3).

Channel side processors (211, 212,
221 and 222) (here, the channel-side processor 221) refers to the I / O request execution state area of the I / O job execution control information 800 (FIG. 4) and sets “2 (Waiting for channel CPU execution) ”is set only when“ 1 (executable) ”is set in the execution possibility field corresponding to the own processor number (221) of the executable processor designation bitmap area. , "3 (executing)" is set in the I / O request execution state area, the corresponding data in the cache memory 232 is transferred to the channel controller CC, and "0 (no execution request) is stored in the I / O request execution state area. ) "Is set. The above is the operation of the data transfer control process at the time of reading. The same applies to writing, and a description thereof will be omitted.

Next, non-stop maintenance processing (control program change processing) of the disk control device 200 will be described. FIG.
FIG. 7 is a main flow diagram of nonstop maintenance processing (control program change processing). The maintenance staff stores a new version of the data transfer control program in the control program storage drives 216 and 226 of the service processors 215 and 225 via the maintenance terminal 240 (step 11).
0). Next, the maintenance person instructs the service processor 215 via the maintenance terminal 240 to start a control program change process (step 120). Service processor 21
5 sets “2 (cluster limited mode)” in the execution cluster instruction information 600 (FIG. 3) (step 125).
Next, the service processor 215 executes a control program change process of the cluster 1 (step 130). The control program change processing of the cluster 1 will be described later in detail with reference to FIG. While the control program change process is being performed on the processors (211, 212, 213, 214) of the cluster 1, the processors (221, 222, 223, 224) of the cluster 2 execute the old version of the data transfer control program. continuing.

Next, the service processor 215 requests the service processor 225 to execute the control program change processing of the cluster 2 (step 140). The service processor 225 executes a control program change process of the cluster 2 (Step 145). The control program changing process of the cluster 2 will be described later in detail with reference to FIG. Cluster 2 processor (221,
222, 223, 224) while the control program change process is being executed, the processor (21
1, 212, 213, 214) are executing a new version of the data transfer control program. When the service processor 215 confirms that the control program change processing of the cluster 2 is completed (step 150), the service processor 215 sets “1 (normal operation mode)” to the operation mode of each cluster in the cluster operation mode instruction information 700 (FIG. 5). Is set and “1” is set in the execution cluster instruction information 600 (FIG. 3).
(Cluster non-limited mode) ”is set (step 15).
5). Then, a message indicating the completion of the control program change process is displayed on the maintenance terminal 240 (step 16).
0).

FIG. 7 is a detailed flowchart of the control program change process (130) of the cluster 1. Cluster 1
Processors (211, 212, 213, 214)
Referring to the executable processor designation bitmap area of the I / O job execution control information 800 (FIG. 4) of all the drives (D0, D1,...), The processor of cluster 1 and the processor of cluster 2 are simultaneously set to “1”. (Executable) ”is confirmed (step 310). Since the execution cluster instruction area of the execution cluster instruction information 600 of FIG. 3 is set to “2 (cluster limited mode)” in step 125 of FIG. 6, the channel-side processor (211, 21)
2,221,222), even if a new access command is passed from the channel control device CC, the executable processor designation bitmap area of the I / O job execution control information 800 (FIG. 4) contains the cluster 1 or Cluster 2
"1 (executable)" is set in only one of the processors. Therefore, the non-cluster limited job will eventually disappear. Confirm this disappearance.

Next, the service processor 215 will be described with reference to FIG.
The operation mode of the cluster 1 in the cluster operation mode instruction information 700 is set to “2 (pause mode)” (step 320). Channel-side processor of cluster 1 (21
1 and 212), when the operation mode of the cluster 1 in the cluster operation mode instruction information 700 is set to “2 (pause mode)”, notify the channel control device CC that a new access command will not be accepted. (Step 320). As a result, a new job to be executed by the processor of the cluster 1 is no longer registered. Next, the processors of cluster 1 (211, 212, 21
3, 214) refers to the executable processor designation bitmap area of the I / O job execution control information 800 (FIG. 4) of all drives (D0, D1,...)
It is confirmed that there is no job (cluster limited job) that is set to "1 (executable)" for the processor (step 325). Then, it notifies the service processor 215 of the completion of the suspension of its own processor (step 330).

Next, the service processor 215 is configured as shown in FIG.
The operation mode of the cluster 1 in the cluster operation mode instruction information 700 is set to “3 (maintenance operation mode)” (step 335). Next, the processor (21) that has detected that the operation mode of the cluster 1 has been set to the maintenance operation mode.
1, 212, 213, 214) request the service processor 215 to transfer a new version of the data transfer control program (step 340). In response, the service processor 215 transfers the new version of the data transfer control program to the processors (211, 212, 213, 214) of the cluster 1 and stores them in each local memory LM (step 345).
Channel-side processor of cluster 1 (211, 212)
Indicates that the new version of the data transfer control program is to be executed on all processors (211, 212,
3, 214), it is notified to the channel control device CC that the reception of a new access command is to be resumed (step 3).
50).

FIG. 8 is a detailed flowchart of the control program change process (145) of the cluster 2 described above. The job existing at the start of the control program change process (145) of the cluster 2 is only the cluster limited job, and there is no non-cluster limited job. First, the service processor 225 checks the cluster operation mode instruction information 70 in FIG.
The operation mode of the cluster 2 of 0 is set to “2 (pause mode)” (step 405). When the operation mode of the cluster 2 in the cluster operation mode instruction information 700 is set to “2 (pause mode)”, the channel-side processors (221 and 222) of the cluster 2 do not accept a new access command. The channel controller CC
(Step 415). As a result, a new job to be executed by the processor of the cluster 2 is no longer registered. Next, the processor of cluster 2 (22
, 222, 223, 224) are all drives (D
0, D1,...), Refer to the executable processor designating bitmap area of the I / O job execution control information 800 (FIG. 4) and set the job (1 (executable)) to the processor of cluster 2 (executable). It is confirmed that there is no cluster-limited job (step 420). Then, it notifies the service processor 225 of the completion of the suspension of its own processor (step 425).

Next, the service processor 225 operates as shown in FIG.
The operation mode of the cluster 2 in the cluster operation mode instruction information 700 is set to “3 (maintenance operation mode)” (step 426). Next, the processor (22) that has detected that the operation mode of the cluster 2 has been set to the maintenance operation mode.
, 222, 223, 224) requests the service processor 225 to transfer the new version of the data transfer control program (step 427). In response, the service processor 225 transfers the new version of the data transfer control program to the processors (221, 222, 223, 224) of the cluster 2 and stores them in each local memory LM (step 430).
Channel-side processors of cluster 2 (221, 222)
Indicates that the new version of the data transfer control program is to be executed on all the processors (221, 222, 22
3, 224), it is notified to the channel control device CC that resumption of acceptance of a new access command is to be resumed (step 4).
40).

According to the above disk controller 200,
Processor group (211, 212, 213, 214, 22
1, 222, 223, and 224) can be upgraded without interrupting the processors that execute different versions of the data transfer control program.

Second Embodiment FIG. 9 is a configuration diagram of a disk control device 200 'according to a second embodiment of the present invention. This disk controller 20
0 'has the same configuration as that of the disk control device 200 of the first embodiment except that the processor group is divided into four clusters.

When the number of cluster divisions is three or more, a plurality of combinations of clusters capable of performing distributed processing of jobs among processor groups can be determined. The combination of clusters that can be distributed is created and stored in the control memory 232 as the cluster combination information 100 as shown in FIG. Cluster Processing Information 1 for Distributed Processing in FIG. 10
At 00, if the processor group having the cluster number on the vertical axis and the processor group having the cluster number on the horizontal axis can perform distributed processing of the job, "OK" is set in a field common to these cluster numbers. If the job cannot be distributed, "NG" is set in a field common to these cluster numbers. Normally, if the processor group with the cluster number on the vertical axis and the processor group with the cluster number on the horizontal axis are executing the same version of the data transfer control program, "OK" is set, and the data transfer control program of the same version is set. If it cannot be executed, "NG" is set. For example, when the processors of all clusters are executing the same version of the data transfer control program, "OK" is set in all fields as shown in FIG. On the other hand, the processor group of cluster 1 (211, 21
Only 3) has started executing the new version of the data transfer control program, and the processors (212, 214, 221, 223, 222,
224), when the execution of the old version of the data transfer control program is continued, "NG" is set in the field corresponding to cluster 1 as shown in FIG. Contains "O
K ”is set. As described above, since the version of the data transfer control program executed by the processor group of each cluster changes during the maintenance operation, the set value of the distributed processable cluster combination information 100 is dynamically changed accordingly. I do.

Each processor has an execution cluster designation area of "2" in the execution cluster designation information 600 of FIG.
(Cluster limited mode) ”, the cluster to which the own processor belongs in the distributed processable cluster combination information 100 in the executable / non-executable field of the executable processor designation bitmap area of the I / O job execution control information 800 in FIG. Is set to "1 (executable)" only in the executable / non-executable field corresponding to the processor belonging to another cluster for which "OK" is set.

According to the above disk control device 200 ', the data transfer control program can be upgraded in finer cluster units. Further, jobs can be distributed among processors having the same version of the data transfer control program.

FIG. 12 is a diagram corresponding to FIG. 11 when the processor group is divided into five clusters. This distributed processable cluster combination information 100 ′ is
And the contents at the time when the program change processing of the processor of the cluster 2 has been completed, the program change processing of the cluster 3 is in the course of being performed, and the program change processing of the processors of the clusters 4 and 5 has not been started yet. I have. That is, since the processors of the clusters 1 and 2 are executing the new program, the clusters 1 and 2 are one set capable of performing the distributed processing. Also, cluster 4 and cluster 5
Is running the old program, clusters 4 and 5 are the other set that can be distributed. Also, since the program of the cluster 3 is being changed, the cluster 3 cannot be distributed with other clusters.

Third Embodiment FIG. 13 is a configuration diagram of a disk control device 200 ″ according to a third embodiment of the present invention. This disk control device 20
0 "is the processor (211, 212, 213, 21
4), the service processor (215) and the shared memory (2)
30) are connected by a bus 201. In this case, the processors (211, 21) connected to the bus 201
2,213,214) into appropriate clusters,
The data transfer control program can be upgraded by the same maintenance method as in the first and second embodiments.

[0036]

According to the method for changing a non-stop program of a multiprocessor system according to the present invention, a program can be changed without stopping the operation of the multiprocessor system entirely and without interfering with programs of different versions. Can be done.

According to the multiprocessor system of the present invention, it is not necessary to stop the operation in order to change the program.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a disk control device according to a first embodiment of the present invention.

FIG. 2 is a configuration diagram of cluster assignment information.

FIG. 3 is a configuration diagram of execution cluster instruction information;

FIG. 4 is a configuration diagram of I / O job execution control information.

FIG. 5 is a configuration diagram of cluster operation mode instruction information.

FIG. 6 is a main flowchart of a non-stop maintenance process (control program change process).

FIG. 7 is a detailed flowchart of a control program change process of the cluster 1;

FIG. 8 is a detailed flowchart of a control program change process of the cluster 2;

FIG. 9 is a configuration diagram of a disk control device according to a second embodiment of the present invention.

FIG. 10 is a configuration diagram of cluster processing information that can be distributed;

FIG. 11 is another configuration diagram of the cluster combination information that can be distributed.

FIG. 12 is a diagram showing still another configuration of the cluster combination information that can be distributed;

FIG. 13 is a configuration diagram of a disk control device according to a third embodiment of the present invention.

[Explanation of symbols]

100, 100 ′ Distributed processing-capable cluster combination information 200, 200 ′, 200 ″ Disk controller 201 Bus 211, 212, 221, 222 Channel side processor 213, 214, 223, 224 Drive side processor 215, 225 Service processor 230 Shared memory 232 Control memory 240 Maintenance terminal 500 Cluster allocation information 600 Execution cluster instruction information 800 I / O job execution control information 700 Cluster operation mode instruction information CC Channel controller D0, D1 Disk drive

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06F 15/16 640 G06F 9/06 630C

Claims (6)

[Claims] 1. A multiprocessor system in which a program is stored in a local memory of each of a plurality of processors and each processor executes a program stored in its own local memory to process a job. From cluster N (N
Is a cluster-limited mode in which N is divided into groups of 2 or more and a natural number equal to or less than the total number of processors, and the job is distributed only among the processors belonging to the same cluster as the cluster to which the own processor belongs. And a cluster non-restricted mode in which jobs are distributed among all processors. During normal operation, jobs are distributed in the cluster non-restricted mode. From the processing of the current job to the processing of the job in the cluster limited mode. When the migration is completed, one cluster is selected, and the processing is performed by the processor group of the non-selected cluster and the selection is continued while accepting a new job. Stop accepting new jobs on the processors in the cluster,
When there are no more jobs to be processed by the processors in the selected cluster, change the programs stored in the local memories of the processors in the selected cluster to new programs, and complete the program change for all the processors in the selected cluster. After accepting a new job and resuming job processing in the processor group of the selected cluster and repeating this for all clusters in order,
A non-stop program changing method for a multiprocessor system, wherein the processing of a job in the cluster-only mode is returned to the processing of a job in the non-cluster-only mode to return to a normal operation. 2. A multiprocessor system in which a program is stored in each local memory of a plurality of processors and each processor executes a program stored in its own local memory to process a job. From cluster N (N
Is grouped into N clusters of 2 or more and a natural number equal to or less than the total number of processors, and the cluster group is a first cluster.
A cluster limited mode in which a job is distributed only to a group of processors belonging to the same cluster as a cluster to which the own processor belongs, and a cluster limited mode in which the job is distributed only to a group of processors belonging to the same cluster as the cluster to which the own processor belongs; A cluster non-restricted mode in which a job is distributed between a processor group belonging to the same cluster and a cluster to which the own processor belongs and a processor group belonging to the cluster belonging to the same group is provided. Clusters belong to one of the first set or the second set, and the jobs are distributed in the non-cluster limited mode. At the time of the program change operation, first, one cluster is selected, and the cluster non-limited mode in the non-selected cluster is selected. Jobs in non-cluster-only mode in the selected cluster while continuing distributed processing of jobs Shifts the process to the job processing in the cluster limited mode, stops accepting new jobs in the processor group of the selected cluster when the shift is completed, and selects the selected cluster when there are no more jobs to be processed in the processor group of the selected cluster. The program stored in each local memory of the processor group of each of the clusters is changed to a new program, and when the change of the program of all the processors of the selected cluster is completed, the reception of a new job and the job The processing is resumed, and the affiliation of the selected cluster is changed from one of the first set or the second set to the other. Then, the processing of the job in the cluster limited mode in the selected cluster is returned to the processing of the job in the non-cluster limited mode. This is repeated for all clusters in order to return to normal operation. And a non-stop program changing method for a multiprocessor system. 3. A multiprocessor system in which a program is stored in each local memory of a plurality of processors and each processor executes a program stored in its own local memory to process a job. From cluster N (N
Is a cluster allocating means for grouping into N clusters up to 2 and a natural number equal to or less than the total number of processors, and a cluster for performing job distribution processing only among processors belonging to the same cluster as the cluster to which the own processor belongs. Distributed processing mode control means for controlling distributed processing in one of a limited mode and a cluster non-limited mode for performing distributed processing of jobs among all processors; In the case of the program change operation, first, the processing is shifted from the processing of the job in the non-cluster limited mode to the processing of the job in the cluster limited mode, and when the transfer is completed, one cluster is selected. Continue processing jobs and accepting new jobs in the processors of the unselected cluster. Stop accepting a new job in the processor group of the selected cluster, and when there are no more jobs to be processed in the processor group of the selected cluster, the programs stored in each local memory of the processor group of the selected cluster are each converted to a new program. After changing the program of all the processors of the selected cluster, the process of accepting a new job and the processing of the job in the processor group of the selected cluster is restarted, and this is repeated for all the clusters in order. A multi-processor system comprising: a program change processing unit that shifts the job processing in (1) to the job processing in the non-cluster mode and returns to normal operation. 4. A multiprocessor system in which a program is stored in each local memory of a plurality of processors and each processor executes a program stored in its own local memory to process a job. From cluster N (N
Is a cluster assigning means for grouping into N clusters of 2 or more and a natural number equal to or less than the total number of processors, and a cluster capable of distributed processing for dividing a cluster group into two sets of a first set and a second set The combination means and a cluster limited mode in which the job is distributed only among the processors belonging to the same cluster as the cluster to which the own processor belongs; or a processor group to which the own processor belongs and a cluster to which the own processor belongs Distributed processing mode control means for controlling distributed processing in one of a cluster non-restricted mode for performing distributed processing of a job among processors belonging to a cluster belonging to the same group as a belonging group, and a normal operation mode Assigns all clusters to one of the first set or the second set, and During the program change operation, first, one cluster is selected, and the job is distributed in the non-cluster limited mode in the non-selected cluster while the distributed processing of the job is performed in the non-cluster limited mode in the non-selected cluster. Shift the job processing to the job processing in the cluster limited mode, and when the shift is completed, stop accepting new jobs in the processor group of the selected cluster, and when there is no job to be processed by the processor group of the selected cluster The program stored in each local memory of the processor group of the selected cluster is changed to a new program, and when the change of the program of all the processors of the selected cluster is completed, the reception of a new job in the processor group of the selected cluster and Resumes job processing and After changing the affiliation of the data from one of the first set or the second set to the other, the processing of the job in the cluster limited mode in the selected cluster is shifted to the processing of the job in the non-cluster limited mode. A multiprocessor system, comprising: a program change processing unit that sequentially returns to a normal operation for a cluster. 5. A multiprocessor system in which a program is stored in each local memory of a plurality of processors and each processor executes a program stored in its own local memory to process a job. Cluster assignment information storage means for storing cluster assignment information indicating which cluster each processor belongs to in order to group into clusters, and a normal operation mode in which the processor processes jobs according to a program stored in a local memory The cluster unit is used to determine whether to use the sleep mode, in which the processor suspends job processing in order to close the cluster, or the maintenance operation mode, in which a program stored in local memory is changed to a new program. The class shown by Operation mode instruction information storage means for storing operation mode instruction information, and a cluster limited mode in which jobs are distributed only to processors belonging to the same cluster as the own processor, or a job between arbitrary processors regardless of the cluster. Cluster instruction information storage means for storing execution cluster instruction information indicating which job is to be processed in the non-cluster limited mode or the distributed processing mode for performing distributed processing, and a program change start instruction to start program change Instructing means, when a program change start is instructed, sets a distributed processing mode to a cluster limited mode, selects an arbitrary cluster, sets an operation mode of the cluster to a pause mode, and sets a processor belonging to the cluster The operation mode of the cluster is Mode is set to the maintenance operation mode, the program stored in the local memory of all the processors belonging to the cluster is changed to a new program, and this is repeated for all the clusters. And a program change processing unit for setting a normal operation mode and a distributed processing mode to a non-cluster limited mode. 6. A multiprocessor system in which a program is stored in each local memory of a plurality of processors and each processor executes a program stored in its own local memory to process a job. Cluster assignment information storage means for storing cluster assignment information indicating which cluster each processor belongs to in order to group into clusters, and a normal operation mode in which the processor processes jobs according to a program stored in a local memory The cluster unit is used to determine whether to use the sleep mode, in which the processor suspends job processing in order to close the cluster, or the maintenance operation mode, in which a program stored in local memory is changed to a new program. The class shown by Operation mode instruction information storage means for storing operation mode instruction information, a cluster limited mode in which jobs are distributed only to processors belonging to the same cluster as the cluster to which the own processor belongs, or belonging to the same cluster as the cluster to which the own processor belongs Execution cluster instruction information indicating whether the job is to be processed in a non-cluster limited mode in which the job is distributed by a processor in another cluster capable of performing the distributed processing of the job between the processor and the own processor, or which distributed processing mode Cluster instruction information storage means for storing the program, program change start instruction means for instructing the start of program change, and when a program change start is instructed, one arbitrary cluster is selected and the distributed processing mode of the cluster is selected. Set to cluster-only mode and The operation mode is set to the sleep mode, and when all the processors belonging to the cluster are in the sleep state, the operation mode of the cluster is set to the maintenance operation mode and stored in the local memory of all the processors belonging to the cluster. Program that changes the existing program to a new program, sets the distributed processing mode of the cluster to the non-cluster limited mode, repeats this for all clusters, and sets the operation mode of all clusters to the normal operation mode at the time of completion. And a change processing unit.