JP3023316B2 - Resource management apparatus for information processing apparatus and resource management method in information processing system - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a resource management method and apparatus suitable for an information processing system that supports multitasking, and more particularly to a method for requesting a memory allocation exceeding a free memory capacity by a program. Information that realizes appropriate and efficient resource management by, for example, selecting a program in the system according to a criterion specified by a system administrator, releasing memory allocated to the program and continuing execution of the program, and the like. The present invention relates to a resource management device in a processing device and a resource management method in an information processing system.

[0002]

2. Description of the Related Art In recent years, the spread of information processing apparatuses has been remarkable, and the amount of information to be processed has been increasing. With the increase in the amount of information, various resource management methods for managing the allocation of the main memory and the secondary storage device have been developed.

Conventionally, even when a relatively high-priority program newly requests the allocation of main memory, if a shortage of main memory occurs, an allocation failure is returned as a result. Or waiting for the program to become available. In particular, when the main memory is insufficient at the time of starting the program, the program that issued the request has to be terminated.

Similarly, when a file is newly created or added, if a file area shortage occurs as a result of requesting a new file area allocation on the disk, an allocation failure is returned as a result. Alternatively, the problem has been dealt with by waiting until it can be assigned.

[0005]

As described above, the conventional problems are as follows. For example, when a program requests a new main memory allocation and the main memory becomes insufficient, it returns an allocation failure as a result or waits until the program becomes available. Since the countermeasures were taken, there was a problem that the execution of the important program was waited for a certain period or could not be started because the less important program was using the main memory.

Similarly, when a new file area is requested on the disk when a new file area is requested and a file area shortage occurs, a failure of allocation is returned as a result. Alternatively, the problem has been dealt with by waiting until the file can be allocated, so that an important program is terminated halfway due to a shortage of a file storage area. Therefore, the present invention has been made in view of the above circumstances, and provides a resource management device of an information processing device and a resource management method of an information processing system that realize appropriate and efficient resource management with simple control. The purpose is to:

[0007]

In order to achieve the above object, the present invention periodically scans a main memory and each page frame constituting the main memory by a predetermined number. A page-out daemon process that goes to sleep after releasing a page frame that has not been accessed by the processor since the previous scan. Means for dispatching at predetermined intervals, means for acquiring the number of free page frames in the main memory, and a page to be scanned by the page-out daemon process every dispatch based on the acquired number of free page frames. Means for setting the number of frames. The management apparatus.

According to the above configuration, when the free space in the main memory is reduced due to, for example, an increase in the number of running programs, the number of page frames scanned by the page-out daemon process each time it is started is determined. On the other hand, if the free space in the main memory increases due to a decrease in the number of running programs or the like, the number of page frames scanned by the page-out daemon process at startup is set to a small value.

When this page-out daemon process is activated, it scans the page frames in the main memory by circulating them by a preset number of times. At this scanning time, the page-out daemon process is provided, for example, corresponding to each page frame. Clear the reference flag specified. This reference flag is set when the processor accesses the page frame. Therefore, the next time the page-out daemon process scans, a page frame in which this reference flag is cleared indicates that no access is being made by the processor during that time. The daemon process supports the effective use of the main memory by releasing the page frame.

That is, according to the resource management apparatus of the present invention, the rate at which the page-out daemon process scans page frames in the main memory can be controlled according to the operating state of the system, and the free space of the main memory becomes tight. If this is the case, the recovery of page frames is promoted by shortening the period during which the page frames can be held without access from the processor. By avoiding excessive release of the used page frame, a flexible operation such as improving the processing performance of the entire system is realized.

[0011]

[0012]

[0013]

[0014]

[0015]

[0016]

[0017]

[0018]

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. (First Embodiment) FIG. 1 is a functional block diagram showing a schematic configuration of a resource management device according to a first embodiment of the present invention.

The overall operation of the resource management apparatus of this embodiment is controlled by the operating system 10.
In the operating system 10, the interval timer 11 generates an interrupt at a predetermined interval, and the clock processing routine 16 is started by the interrupt, and the clock processing routine 16 starts the page-out daemon process 20 each time.

This page-out daemon process 20
When the program is started, it traverses a page frame table 29 indicating the state of each page frame in the main memory 30 by a preset number (when the last page frame is reached, returns to the first page frame). In this case, scanning is performed sequentially. At this time, reference bits (reference bi) provided corresponding to each page frame are provided.
t) is checked, and when the reference bit is set (1), the reference bit is cleared (0).

This reference bit is set when a processor (not shown) accesses the page frame. Therefore, when the reference bit is inspected, the reference bit is cleared. If there is, the access by the processor has not been performed since the previous scanning (while the page-out daemon process 20 makes one round of scanning). Therefore, in this case, the page-out daemon process 20 releases this page frame. As a result, a page frame that has not been accessed for more than a predetermined interval does not continue to exist in the main memory 30, and effective use of the main memory 30 is realized.

The main memory management unit 12 of the present embodiment
The number of used page frames and the number of empty page frames of the main memory 30 are written in the memory management information 13. Then, the page-out daemon process 20 determines the number of page frames to be scanned based on the information on the main memory 30 acquired from the memory management information 13. Then, the page-out daemon process 20 sleeps until a next interval timer interrupt occurs after scanning the predetermined number of page frames. Also,
As another method, the number of page frames to be scanned by the main memory management unit 12 can be determined, and the value can be passed to the page-out daemon process 20.

If the amount of free memory in the main memory 30 is small, the scanning rate of the page-out daemon process 20 is changed by setting a large number of page frames to be scanned by the page-out daemon process 20 at one startup. In the case where there is enough free memory in the main memory 30, the number of page frames scanned by the page-out daemon process 20 at one startup is reduced, and the number of used page frames is reduced. By avoiding excessive recovery of the system, the throughput of the entire system is improved.

As a result, appropriate and efficient resource management according to the operating status of the system is performed. The same effect can be obtained by adjusting the frequency at which the clock processing routine 16 activates the page-out daemon 20 instead of changing the number of scanned page frames of the page-out daemon process 20 to be variable. It is.

(Second Embodiment) Next, a second embodiment of the present invention will be described with reference to FIGS. FIG. 2 shows the second
It is a functional block diagram showing a schematic structure of a resource management device concerning an embodiment.

The overall operation of the resource management apparatus of this embodiment is controlled by an operating system 10, which manages memory management information 13 and process management information 14. Then, when the amount of free memory managed by the memory management information 13 falls below a predetermined value, the operating system 10 sets the main memory release process 40 to an executable state (schedule waiting state).

The main memory release process 40 operates while referring to the memory management information 13 and the process management information 14 managed by the operating system 10. As shown in FIG. 2, the main memory release process 40 first executes an empty area shortage waiting step 41. In this step, a free space shortage system call 10a in the operating system 10 is called, and a request is made to sleep this process until a free space shortage in the main memory 30 occurs.

Then, it is assumed that an allocation of the main memory 30 is newly requested from one of the programs, and that there is no free area enough to meet the request, that is, that a memory shortage has occurred. At this time, the operating system 10 makes the sleeping main memory release process 40 executable. Eventually, the main memory release process 40 that has become executable becomes the process management information 1 in the target process selection step 42.
4, a process with a low priority is selected, for example. Then, a memory release step 43 releases the memory allocated to this process. As means of this release, abort of the process or swap-out can be considered. With this release, the memory shortage is resolved, the allocation is made to the requesting process, and the processing is continued.

As a result, for example, since a low-priority process is present, a disadvantage that a high-priority process is not executed for a long time is eliminated, and appropriate resource management is performed.

In the case where, for example, a process that is started periodically is resident, it is effective to select a process that has a long time until the next execution time instead of the priority. Yes, it is also effective to select a batch process that does not specify an end time, and it is also effective to select a process with less updated memory in consideration of the processing at the time of releasing memory. It is. These can be realized by an information management method in the process management information 14. Further, for example, releasing the memory allocated to the main memory release process 40 itself is also effective in some cases.

The main memory release process 40 is started only when the free space of the main memory 30 falls below a predetermined value. No overhead occurs. The operating system 10
Since it is constructed as an application program that operates in the special mode under the application program, the resource management rules of the main memory 30 can be added and changed only by rewriting this application program, and its expandability can be secured.

FIG. 3 shows an operation procedure of the resource management device of the present embodiment. In the resource management device of the present embodiment,
First, a menu screen for setting a resource management rule of the main memory 30 is displayed (step A1). On this screen, for example, as shown in FIG. 4A, a list of processing methods when the main memory is insufficient is displayed to prompt the system administrator to make a selection. Then, the system administrator inputs the number of the desired processing method from the presented processing methods into the field 51 indicated by the prompt (step A2).
FIG. 4B shows the selection result of the system administrator. In this case, processing method 2, processing method 1, processing method 4
The process to release the memory is determined in this order.

When this input is made, the resource management device performs pre-processing of the specified processing method, and prepares the main memory release process 40 to operate in the specified processing method (step A3). .

Thereafter, the main memory release process 40
It enters a state of waiting for free space (step A4). When the free space of the main memory 30 falls below a predetermined value, the operating system 10 activates the main memory release process 40, and the main memory release process 40 selects a target process (step A).
5), the main memory is released (step A6). These steps A4 to A6 are executed by the main memory release process 40 from the operating system 10.
Is repeated until an instruction to end is given.

Note that, in the target process selection step of step A5, the field 51 in FIG.
Each processing method will be adopted according to the priority specified in. That is, field 5 shown in FIG.
In 1, the numbers are designated, for example, in the order of 2, 1, and 4. Therefore, first, if there is a process corresponding to “2”, the process is selected, and if not, if there is a process corresponding to “1”, it is selected, and so on. Then, the used main memory is released by the selected process.

Next, the operation principle of the resource management device is shown in FIG.
This will be described with reference to FIG. Now, the main memory 30 is
Are assigned as shown on the left side of. Here, assuming that the application program A newly requests the allocation of the main memory 30 exceeding the free space, the main memory release process 40 forcibly removes the memory used by the application program B having the lowest priority. release. Then, the released memory area is allocated to the application program A, and the execution of the application program A having a high priority is continued as shown on the right side of FIG.

Thus, appropriate resource management is realized. (Third Embodiment) Next, a third embodiment of the present invention will be described with reference to FIGS.

FIG. 6 is a functional block diagram showing a schematic configuration of the resource management device according to the third embodiment. The resource management device of the present embodiment is entirely controlled by an operating system 10, which manages file management information 15. Then, when the free space of the disk 70 managed by the file management information 15 falls below a predetermined value, the operating system 10 sets the file compression process 60 in an executable state.

This file compression process 60 is performed by the file management information 1 managed by the operating system 10.
5 and operates. File compression process 60
As shown in FIG.
1. It comprises a file selection step 62 and a file compression step 63. First, an empty area shortage waiting step 61
Calls the operating system 10 and sleeps until a state in which the free space of the disk 70 becomes insufficient appears.

Now, a new creation or addition of any file is newly requested, and as a result of responding to this request, the remaining free space is less than the limit value, and there is no free space to meet the demand. Assume that an area shortage has occurred. At this time, the operating system 10
Dispatching the file compression process 60, referring to the file management information 15 in the file selection step 62,
For example, select the file with the oldest reference date. Then, in a file compression step 63, the file is compressed. Therefore, when the shortage of the file area is resolved by this compression, the requested file is newly created or added, and the process is continued.

As a result, more efficient resource management is performed. It is also effective to select, for example, a file that exists as a backup. These can be realized by the information management method in the file management information 15. Furthermore, for example, the size of the file to be compressed is predicted from the size of the file area that must be extracted, and a file having a size and type that is predicted to have higher compression efficiency can be selected by selecting a file that is close to that size. Selecting a file is also effective. For example, MPEG format or JP
It can be predicted that the document file has higher compression efficiency than the EG format file.

This file compression process 60 also
Since it is started when the free space of the disk 70 falls below a predetermined value, unnecessary overhead does not occur when the free space of the disk 70 has room. Further, since the application program is constructed as an application program that operates under the operating system 10, the resource management rules of the disk 70 can be added and changed only by rewriting the application program, and its expandability can be secured.

FIG. 7 shows an operation procedure of the resource management apparatus according to the present embodiment. In the resource management device of the present embodiment,
First, a menu screen for setting a resource management rule for the disk 70 is displayed (step B1). On this screen, similarly to the list display shown in FIG. 4A, for example, a list of processing methods when the free space of the disk 70 is insufficient is displayed.
Prompt the system administrator for a choice. Then, the system administrator inputs the number of the desired processing method from the presented processing methods in the field indicated by the prompt (step B2).

When this input is made, the resource management device performs preprocessing of the designated processing method, and prepares the file compression process 60 to operate in the designated processing method (step B3).

Thereafter, the file compression process 60 waits for free space (step B4). Disk 70
When the free space of the file falls below a predetermined value, the file compression process 60 is executed by the operating system 10, and the file selection (step B5) and the file compression (step A6) are performed by the file compression process 60. Become. These steps B4 to
B6 is repeated until the operating system 10 gives an instruction to end the main memory release process 40.

In the file selection step of step B5, each processing method is adopted according to the priority specified in the field on the menu screen as in the second embodiment. Examples of one of the plurality of processing methods in the present embodiment include “select the file with the oldest update date” and “select the backup file with priority”.
The selection of each processing method is performed according to the procedure described in the second embodiment. Then, the selected file is to be compressed.

Here, the operation principle of the present embodiment will be described with reference to FIG. Now, it is assumed that the disk 70 is allocated as shown on the left side of FIG. Here, if the file A newly requests the allocation of a disk exceeding the free space, the file compression process 60 executes the predetermined processing method (a plurality of processing methods input in the fields on the menu screen). For example, by selecting the file with the oldest update date), the file B is selected and compressed. Consequently freed file regions is will be assigned as an additional component of the file A, Figure 8
The execution will be continued as shown on the right side of FIG.

As a result, appropriate resource management is performed. For example, as shown in FIG. 9, a file existing as a backup (with an extension of “.b
If there is a file with "ak" added), selecting these files preferentially realizes more appropriate resource management.

Further, as shown in FIG. 10, for example, as a result of performing a compression process, by selecting a file which is estimated to have a free space which is estimated to be the closest to the file capacity to be extracted, compression can be performed. C spent
PU time can be saved.

In the flowchart of FIG. 7, the file is selected and compressed when the free space of the file falls below a predetermined value (first value). Free space is a predetermined value (second value)
The file may be selected and decompressed when the data is recovered up to. FIG. 11A and FIG.
1B. FIG. 11A relates to file compression processing, and FIG. 11B relates to file expansion.

First, the procedure of the file compression process will be described. Now, by referring to the file management information 15 in the operating system 10, the disk 7
The file free space amount within 0 is monitored (step C1). Then, it is determined whether or not the free space amount obtained from the file management information 15 is smaller than a predetermined value L1 (step C2). If it is not smaller than L1, the file compression process is not performed, and if it is smaller than L1, the operating system 10
Dispatch. Thus, in the file compression process 60, a file to be compressed is selected based on the processing method determined in the above-described procedure (step C).
3). Next, the selected file is compressed (step C4).

Next, the procedure of the file decompression process will be described. As in the above case, the operating system 10
, The amount of free file space in the disk 70 is monitored by referring to the file management information 15 (step D1). Then, it is determined whether the free space amount obtained from the file management information 15 is larger than a predetermined value L2 (step D2). If it is not larger than L2, the file decompression process is not performed. If it is larger than L2, the operating system 10 dispatches a file decompression process (not shown). Thus, in the file decompression process, a file to be decompressed is selected based on a predetermined processing method (step D).
3). Next, the selected file is expanded (step D4).

By adopting the decompression process as described above, the file can be efficiently accessed when there is a margin in the file capacity. In the above embodiment, the file compression program is described as first waiting in the free space shortage step until the file area becomes insufficient. However, the same applies to the file expansion process in the operating system as another method. Further, the file compression program may be started when the shortage of the file area is detected.

(Fourth Embodiment) Next, a fourth embodiment of the present invention will be described with reference to FIG. FIG. 12 is a functional block diagram illustrating a schematic configuration of the resource management device according to the fourth embodiment.

The resource management apparatus according to the present embodiment has a link creation step 64 and a file movement step in the file compression process 60 in addition to the configuration of the resource management apparatus according to the third embodiment.
And a step 65.

[0057] In the file compression process 60 of the present embodiment, for example when the pressure <br/> condensation of files of a disk 70a which is selected by the file selecting step 62 with file compression step 63, the file moving step 65, the compressed file The disk is moved from the disk 70a to the disk 70b, which is a subdisk, and the link
Then , this link information is created and stored in the source of the disk 70a. Thereby, the resource management of the disk 70a as the main disk is performed more appropriately.

[0058]

As described in detail above, according to the present invention, a cycle in which the page-out daemon process scans the main memory is controlled in accordance with the operation status of the system, so that appropriate and efficient resource management is realized. Is done.

When an area shortage occurs in the main memory, an appropriate program is selected and a page frame in the main memory allocated to the program is released, so that a less important program can be stored in the system. To prevent the execution of important programs.

Further, even when an area shortage occurs in the disk device, an appropriate file is selected and a free area is obtained by compressing this file. This realizes more efficient resource management.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing a schematic configuration of a resource management device according to a first embodiment of the present invention.

FIG. 2 is a functional block diagram showing a schematic configuration of a resource management device according to the embodiment;

FIG. 3 is a flowchart illustrating an operation procedure of a resource management device according to a second embodiment of the present invention.

FIG. 4 is an exemplary view showing a screen display for prompting a system administrator to select a processing method when the main memory is insufficient in the resource management apparatus according to the embodiment.

FIG. 5 is a conceptual diagram illustrating the operation principle of the resource management device according to the embodiment.

FIG. 6 is a functional block diagram illustrating a schematic configuration of a resource management device according to a third embodiment of the present invention.

FIG. 7 is an exemplary flowchart for explaining the operation procedure of the resource management device according to the embodiment;

FIG. 8 is a conceptual diagram illustrating the operation principle of the resource management device according to the embodiment.

FIG. 9 is a conceptual diagram illustrating the operation principle of the resource management device according to the embodiment.

FIG. 10 is a conceptual diagram illustrating the operation principle of the resource management device according to the embodiment.

FIG. 11 is an exemplary flowchart for explaining the operation procedure of the resource management apparatus according to the embodiment;

FIG. 12 is a functional block diagram showing a schematic configuration of a resource management device according to a fourth embodiment of the present invention.

[Explanation of symbols]

10 operating system, 12 main memory management unit, 13 memory management information, 20 page-out daemon process, 29 page frame table, 30
... Main memory, 40 ... Main memory release process, 6
0: File compression process, 70: Disk.

Continuation of the front page (56) References JP-A-57-169987 (JP, A) JP-A-59-207479 (JP, A) JP-A-5-216744 (JP, A) JP-A-5-158744 (JP) , A) (58) Field surveyed (Int. Cl. ⁷ , DB name) G06F 9/46 G06F 12/12 G06F 12/00 G06F 12/02 G06F 9/06

Claims (4)

(57) [Claims] 1. A main memory and each page frame constituting the main memory are scanned periodically by a predetermined number of times, and page frames that have not been accessed by the processor since the previous scan are scanned. A resource management device for use in an information processing system including a page-out daemon process that goes into a sleep state after being released, comprising: a unit that dispatches the page-out daemon process at a predetermined interval; Means for acquiring the number of empty page frames; and means for setting the number of page frames to be scanned by the page-out daemon process each time dispatching, based on the acquired number of empty page frames. Resource management device for information processing equipment. 2. The information processing apparatus according to claim 1, wherein said setting means includes means for setting the number of page frames scanned by the previous page out daemon process to increase as the number of free page frames decreases. Resource management device. 3. The resource management apparatus according to claim 1, wherein said setting means includes means for shortening an interval for dispatching a page-out daemon process as the number of free page frames decreases. 4. A main memory and each page frame in the main memory are scanned in a predetermined number of cycles, and after releasing a page frame that has not been accessed by the processor since the previous scan, A resource management method in an information processing system including a page-out daemon process that is in a sleep state, wherein the number of empty page frames in the main memory is acquired, and the page is determined based on the acquired number of empty page frames. A resource management method in an information processing system, wherein the number of page frames to be scanned by an out-daemon process at each dispatch is set.