JP2008225818A - Information processor, process compression method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an information processor for improving performance by suppressing mounting costs and memory usage. <P>SOLUTION: This information processor for performing a predetermined process on a main memory is provided with a process compression means for compressing the image of a process being executed on a main memory based on a compression rule corresponding to a process being executed. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an information processing apparatus, and more particularly to an information processing apparatus that performs process compression.

  Conventionally, when processing is performed for each client in a network file server or the like, an implementation method is often used in which a process for performing the response is started on the server side for each communication path with the client. In such a mounting method, because of the independence of individual processes, it is not necessary to consider the process of other response, so the development difficulty level is low and the development efficiency is good. On the other hand, the amount of data held by each process tends to increase, and if there are tens of processes, there is no problem, but in the case of a server that requires thousands of processes, the memory space Holding becomes difficult.

  In many cases, swap processing for a secondary storage device is used as a process compression processing method. For example, Patent Document 1 discloses a technique of compressing a process image to reduce its size and swapping it out to a secondary storage device.

JP 2000-276363 A

  However, the compression processing using swap as described above causes degradation of the performance of the apparatus, particularly degradation of response time. In other words, when restoring a process swapped out to the secondary storage device, it is necessary to read from the secondary storage device that is relatively slow and execute the restoration process, resulting in a delay in the process. On the other hand, in order to solve the above problems, there are methods such as reducing the data space as much as possible and managing shared data between processes, but the development difficulty increases, and the development efficiency and failure rate also increase. There was a problem of going up.

  For this reason, the present invention provides an information processing apparatus that improves the disadvantages of the above-described conventional example, and in particular, can suppress the mounting cost and suppress the memory usage and improve the performance. For that purpose.

  Accordingly, an aspect of the present invention is an information processing apparatus that executes a predetermined process on a main memory, and an image of the process being executed on the main memory is based on a compression rule corresponding to the process being executed. It is characterized by having a process compression means for compressing.

  According to the above invention, first, when a process is being executed on the main memory, it is necessary to compress the process being executed in a situation where the memory area is tight. In such a case, the process is compressed using a compression rule corresponding to the process being executed. Thereby, compression with a high compression rate can be realized according to the characteristics of the process, and the main memory can be used efficiently. Furthermore, since the compression ratio is high, even if the compressed data is left in the main memory, the memory capacity is not reduced and the process can be returned at high speed.

  In addition, the information processing apparatus includes a compression dictionary storage unit that stores compression dictionary data representing a compression rule for each process, a compression dictionary selection unit that selects compression dictionary data corresponding to an ongoing process to be compressed, The process compression means compresses the image of the process being executed using the selected compression dictionary data. The information processing apparatus further includes a compression dictionary generation unit that generates compression dictionary data representing a compression rule corresponding to each process being executed, and the process compression unit uses the generated compression dictionary data. It is characterized by compressing the image of a running process.

  Thus, the compression process is executed using the compression dictionary data prepared in advance according to the process being executed or the compression dictionary data generated according to the process being executed before the compression process. Therefore, the compression process can be executed more efficiently and at a high speed.

  The information processing apparatus includes a compressed data storage means for storing the compressed data compressed by the process compression means in a predetermined storage device. The compressed data storage means stores the compressed data in a predetermined auxiliary storage device. In particular, the compressed data storage means stores the compressed data in the main memory device when the compressed data can be compressed to a predetermined ratio or less compared with the image before compression, and otherwise the compressed data is stored in the predetermined data. It is characterized in that it is stored in an auxiliary storage device.

  Thereby, the main memory can be efficiently used by storing the compressed data obtained by compressing the process as described above in a predetermined storage device such as an auxiliary storage device. In particular, when the compression rate is high, even if the compressed data is left in the main memory, the process can be restored by reading the compressed data at a high speed while suppressing the used capacity of the main memory.

  In the present invention, the compression rule is set according to the type of the process being executed. Further, the compression rule is set according to a virtual address set for a process being executed. In addition, the above process is characterized in that it is a process having a part of information unique to each process.

  Thereby, it can compress by the compression rule according to the characteristic for each process, and can implement more efficient compression. In particular, for the same type of process having unique information in the process, it is only necessary to perform compression so that at least the unique information portion is included, so that more efficient compression can be realized.

  Another embodiment of the present invention is a process compression method in which an information processing apparatus compresses a predetermined process executed on a main memory, and an image of a process being executed on the main memory is It has the process compression process which compresses based on the compression rule according to a process, It is characterized by the above-mentioned.

  In the process compression method, the information processing apparatus stores in advance compression dictionary data representing a compression rule for each process, and a compression dictionary selection step of selecting compression dictionary data corresponding to a running process to be compressed And the process compression step compresses the image of the process being executed using the selected compression dictionary data. The process compression method further includes a compression dictionary generation step for generating compression dictionary data representing a compression rule corresponding to each process being executed, and the process compression step includes generating the generated compression dictionary data. It is characterized by using it to compress the image of a running process.

  Further, the method includes a compressed data storage step of storing the compressed data compressed in the process compression step in a predetermined storage device. In particular, the compressed data storage step stores the compressed data in the main memory device when the compressed data can be compressed to a predetermined ratio or less compared with the image before compression, and otherwise the compressed data is stored as a predetermined auxiliary. It is stored in a storage device.

  Furthermore, a program according to another aspect of the present invention provides an information processing apparatus that executes a predetermined process on a main memory, and an image of the process being executed on the main memory, and a compression rule corresponding to the process being executed. It is characterized by realizing a process compression means for compressing based on the above.

  In the above program, the information processing apparatus stores in advance compression dictionary data representing compression rules for each process, and the information processing apparatus selects compression dictionary data corresponding to the running process to be compressed. While realizing the dictionary selection means, the process compression means compresses the image of the process being executed using the selected compression dictionary data. Further, the program causes the information processing apparatus to realize a compression dictionary generation unit that generates compression dictionary data representing a compression rule corresponding to each process being executed, and the process compression unit generates the compression dictionary data. It is characterized by compressing an image of a process being executed using compression dictionary data.

  Further, the program is characterized in that the information processing apparatus realizes a compressed data storage means for storing the compressed data compressed by the process compression means in a predetermined storage device. In particular, the compressed data storage means stores the compressed data in the main memory device when the compressed data can be compressed to a predetermined ratio or less compared with the uncompressed image, and in other cases, the compressed data is stored as the predetermined auxiliary data. It is stored in a storage device.

  Even the method and program of the above-described configuration operate in the same manner as the information processing apparatus, and thus the above-described object of the present invention can be achieved.

  Since the present invention is configured and functions as described above, according to this, since compression with a high compression rate can be realized according to the characteristics of the process, efficient use of the main memory can be realized, and compressed data can be mainly used. Even if it is left in the memory, it has an unprecedented excellent effect that the process can be returned at high speed without constraining the memory capacity.

  The present invention is characterized by compressing a process image in accordance with process characteristics. Hereinafter, a specific configuration and operation will be described with reference to examples.

  A first embodiment of the present invention will be described with reference to FIGS. 1 and 2 are functional block diagrams showing the configuration of a computer, and FIG. 3 is a diagram showing a process. 4 to 5 are flowcharts showing the operation of the computer.

[Constitution]
A computer 1 shown in FIGS. 1 and 2 is a server computer such as a file server connected to a network, and is a computer capable of executing a plurality of processes. As its configuration, first, it includes a CPU 10 that is an arithmetic device, a main memory 20 that is a work space for each process, and a hard disk drive 30 (HDD) that is an auxiliary storage device. The auxiliary storage device may be any type of storage device. Here, the computer 1 shown in FIGS. 1 and 2 is the same computer, and is divided into two diagrams for convenience of explanation.

  Then, by incorporating a predetermined program into the CPU 10, as shown in FIG. 1, a process execution processing unit 11, a compression dictionary selection processing unit 12, a compression processing unit 13, a compressed data storage processing unit 14, , Has been built. Further, as shown in FIG. 2, a process restoration processing unit 15 is also constructed in the CPU 10. In the main memory 20, a process 21 to be executed is generated, and a compressed data storage unit 22 for storing compressed data to be described later is formed. Further, the HDD 30 is formed with a compression dictionary storage unit 31 and a compressed data storage unit 32. Hereinafter, each configuration will be described in detail.

  First, the process execution processing unit 11 generates a process 21 in the virtual space in the main memory 20 and executes the process 21. Here, in this embodiment, it is assumed that the process executed by the computer 1 that is a network server is a process generated individually for each client terminal accessed from the network. In particular, the process for each client is the same type of process, and is generated from a copy of a single process. In the process, unique information for each process such as process type, virtual address in process, destination address, have. An example of the process is shown in FIG. As shown in this figure, when process A and a plurality of processes of the same type are generated in the virtual address space on the main memory, specific information for each process is stored in the predetermined virtual address portion of process A. It has become a state.

  Then, when the number of processes being executed increases and the area of the main memory 20 has become tight, the process execution processing unit 11 specifies an unused process as a compression target, and the main memory of the specified process The image (content), process type, and in-process virtual address on 20 are extracted and notified to the compression dictionary selection processing unit 12. As shown in FIG. 2, the process execution processing unit 11 notifies the process restoration processing unit 15 of information for specifying the process when returning the once compressed process.

  The compression dictionary selection processing unit 12 (compression dictionary selection unit) converts the compression dictionary data stored in the compression dictionary storage unit 31 (compression dictionary storage unit) of the HDD 30 according to the process being executed as a compression target. Select. Here, the compression dictionary data stored in the compression dictionary storage unit 31 is data representing a compression rule set in advance corresponding to a process type and / or a virtual address in the process. That is, the compression dictionary is set so as to be highly compressible according to the structure and characteristics of the process. For example, in this embodiment, since the process is generated from a copy of a single process, the configuration of each process is almost the same if it is the same type, and the difference is, for example, the destination address that the process has It is only unique information. Therefore, the stored compressed dictionary data is at least process type, intra-process virtual address, unique, based on virtual address information such as which virtual address is stored in which virtual address in each process. This is dictionary data for generating compressed data so as to hold information and the like in the compressed data. Further, by storing the data common to the processes of the same type in the compressed dictionary data, the capacity of the compressed data can be further reduced.

  The compression dictionary data as described above is prepared in advance, and the compression dictionary selection processing unit 12 responds on the basis of the process type and the in-process virtual address of the process to be compressed notified from the process execution processing unit 11. The compression dictionary data to be searched is retrieved from the compression dictionary storage unit 31 and read out. Then, the compression dictionary data and the process image to be compressed notified from the process execution processing unit 11 are transferred to the compression processing unit 13.

  The compression processing unit 13 (process compression means) receives the compression dictionary data and the process image passed from the compression dictionary selection processing unit 12, compresses the process image using the compression dictionary data, and compresses the compressed data. Execute the process to generate. At this time, since the process image is mostly the same as the process image of the same type, the portion common to the other portions is not included in the compressed data, and as described above, at least the compression type, the in-process virtual address, the unique information To generate compressed data. That is, a common part of the process is included in the compression dictionary data, and the compressed image is generated by compressing the process image before compression so as to include the minimum necessary information that can be restored from the compressed data. Then, the generated compressed data is transferred to the compressed data storage processing unit 14.

  The compressed data storage processing unit 14 (compressed data storage means) stores the generated compressed data in the main memory or the HDD. In this embodiment, the storage destination of the compressed data is set according to the compression rate of the compressed data. For example, if the compressed data capacity can be compressed to less than half of the original process image, the compression ratio is high, so it is considered that the memory area is not so tight and is held in the main memory 20 as it is. deep. That is, it is stored in the compressed data storage unit 22 of the main memory 20. On the other hand, if the compressed data capacity cannot be compressed to less than half of the original process image, it is stored in the compressed data storage unit 32 of the auxiliary storage device (secondary storage device) such as the HDD 30. It should be noted that other values of compression rates may be set as criteria for determining the storage location of compressed data. For example, when compression can be performed to one third or less, the compressed data may be set to be stored in the main memory 20, and can be arbitrarily set according to the capacity of the main memory 20 or the HDD 30.

  The process restoration processing unit 15 is a process for restoring a once compressed process. Specifically, the process notified from the process execution processing unit 11 is specified, and the compressed data of the process is extracted from either the compressed data storage unit 22 of the main memory 20 or the compressed data storage unit 32 of the HDD 30. . Also, compressed dictionary data corresponding to this process is extracted. Then, the compressed data is restored based on the compressed dictionary data. In other words, by incorporating the individual data included in the compressed data into the newly generated process, the specific process that is the return target can be restored.

[Operation]
Next, the process compression operation and the process restoration operation in the computer having the above configuration will be described with reference to the flowcharts of FIGS. FIG. 4 shows the process compression operation, and FIG. 5 shows the process restoration operation.

  First, when a large number of processes 21 are executed on the main memory 20 of the computer 1 and the main memory 20 is tight (step S1), a process of compressing an unused process image is started. . Then, the process to be compressed is specified (step S2), and the process image, process type, and in-process virtual address of the process are extracted (step S3).

  Subsequently, based on the notified process type and in-process virtual address, compression dictionary data corresponding to the feature of the process to be compressed is searched and selected from the compression dictionary storage unit 31 (step S4, Compression dictionary selection process). Then, using this selected compression dictionary data, the process image to be compressed is compressed to generate compressed data (step S5, process compression step).

  If the process image can be compressed to less than half by the compression process (Yes in step S6), the compressed data is stored in the main memory 20 (step S7, compressed data storage step). On the other hand, if the data cannot be compressed to less than half (No in step S6), the compressed data is written and stored in the HDD 20 (step S8, compressed data storage step).

  Thereafter, when the process execution processing unit 11 determines that the already compressed process is necessary (Yes in step S11), the process to be restored is specified, and the compressed data storage location of the process, that is, the main data is stored. It is determined whether it is stored in the memory 20 or whether it is stored in the HDD 30 (step S12). Then, the compressed data is read (step S13), a compression dictionary corresponding to the process is selected (step S14), and the process is restored on the main memory 20 and executed (step S15). If the compressed data is stored in the main memory 20 in the step 13, the process can be restored by reading the compressed data at a high speed.

  As described above, according to the present invention, it is possible to perform compression at a high compression rate using compression dictionary data according to the characteristics of the process. Even if it is stored in the memory, the capacity of the main memory can be saved. Accordingly, swap-out to a secondary storage device such as a hard disk drive having a relatively low data input / output speed can be suppressed, and subsequent processes such as process restoration can be executed at high speed. As a result, the performance of the computer itself can be improved.

  Further, in the present invention, since the configuration such as the compression processing unit 13 as described above may be newly added and mounted between the main memory 20 and the HDD 30, it does not affect other elements of the system. Can be easily realized. As a result, normal memory usage reduction etc. must be done by reviewing each individual program, etc., but implement a function that can realize efficient use of memory at low cost without such review be able to.

  Next, a second embodiment of the present invention will be described with reference to FIGS. FIG. 6 is a functional block diagram showing the configuration of the computer 1 in this embodiment, and FIG. 7 is a flowchart showing its operation.

[Constitution]
As shown in FIG. 6, a computer 1 having a process compression function in the present embodiment has a configuration substantially similar to that of the computer 1 in the first embodiment described above. As a specific difference, in the present embodiment, the compression dictionary data is not stored in the compression dictionary storage unit 31 ′ in advance, and when the computer 1 performs the compression process, the process to be compressed is processed. A compression dictionary generation processing unit 12 ′ (compression dictionary generation means) that generates compression dictionary data suitable for the process based on the characteristics is provided. The compression processing unit 13 is configured to execute the compression process of the process using the generated compression dictionary data.

  Here, the compression dictionary generation processing unit 12 ′, for example, based on the structure, type, in-process address, unique information, etc. of the process image itself that is the compression target notified from the process execution processing unit 11 described above, Optimal compression dictionary data is generated to compress this process. Specifically, based on a virtual address that represents the structure of the process image, process-specific information and a common part in the same type of process are identified, and compressed dictionary data is generated based on the analysis result of the structure. . For example, compression dictionary data is generated that represents a compression rule in which the common part is stored in the compression dictionary data, and an in-process address indicating process type and structure, unique information, and the like are generated so as to be included in the compressed data. Then, the compression dictionary generation processing unit 12 ′ stores the generated compression dictionary data in the compression dictionary storage unit 31 ′ for use in future restoration or compression of the same type of process.

[Operation]
Next, the process compression operation of the computer 1 in the second embodiment will be described with reference to the flowchart of FIG. In addition, since it is generally the same as that of the first embodiment, a brief description will be given.

  First, when a large number of processes 21 are executed on the main memory 20 of the computer 1 and the main memory 20 is tight (step S1), a process of compressing an unused process image is started. . Then, the process to be compressed is specified (step S2), and the process image, process type, and in-process virtual address of the process are extracted (step S3).

  Subsequently, based on the notified process type and in-process virtual address, new compression dictionary data corresponding to the feature of the process to be compressed is generated (step S4 ', compression dictionary generation step). Then, by using the generated compression dictionary data, the process image to be compressed is compressed to generate compressed data (step S5, process compression step). At this time, the used compression dictionary data is stored in the compression dictionary data storage unit 31 '.

  If the process image can be compressed to less than half by the compression process (Yes in step S6), the compressed data is stored in the main memory 20 (step S7, compressed data storage step). On the other hand, if the data cannot be compressed to less than half (No in step S6), the compressed data is written and stored in the HDD 20 (step S8, compressed data storage step).

  Next, a third embodiment of the present invention will be described with reference to FIGS. 8 and 9 are functional block diagrams showing the configuration of the computer 1, respectively.

  First, the computer 1 having the configuration shown in FIG. 8 has a configuration substantially similar to that of the computer 1 of the first embodiment described above. As a specific difference, in this embodiment, an auxiliary memory 130 such as a flash memory is provided in place of the HDD, and a compression dictionary storage unit 131 and a compressed data storage unit 132 are formed therein. ing. The compressed data storage processing unit 14 ′ writes all the compressed data to the auxiliary memory 130 that is a secondary storage device.

  According to such a configuration, since compressed data is not written out to the main memory 20, it is effective for a computer with a small capacity of the main memory 20, for example. In the present invention, since the compression efficiency is high as described above, the volume of compressed data is smaller than the conventional one, and the capacity required for the auxiliary memory 130 can be reduced.

  FIG. 9 shows another example of the computer 1. The computer 1 shown in this figure also has substantially the same configuration as the computer 1 of the first embodiment described above. As a specific difference, in this embodiment, a ROM 230 is provided instead of the HDD, and a compression dictionary storage unit 131 is formed therein. The compressed data storage processing unit 14 ″ writes all the compressed data to the main memory 20.

  According to such a configuration, even in a computer such as an embedded device that is not equipped with an auxiliary storage device such as a flash memory or an HDD, the compressed data written to the main memory 20 has high compression efficiency. The memory can be used efficiently while suppressing the tightness of the memory.

  In the configuration shown in FIGS. 8 and 9, the compression dictionary data may be prepared in advance. As described in the second embodiment, the compression dictionary data is stored based on the structure of the process to be compressed. It may be generated and used for compression processing.

  The present invention can be used in any information processing apparatus such as a network server that executes a large number of processes of the same type, and an embedded device that does not include a secondary storage, and has industrial applicability.

FIG. 3 is a functional block diagram illustrating a configuration of a computer according to the first embodiment. FIG. 3 is a functional block diagram illustrating a configuration of a computer according to the first embodiment. It is a figure which shows the structure of a process. 3 is a flowchart illustrating an operation of a process compression process by a computer in the first embodiment. 3 is a flowchart illustrating an operation of a process restoration process performed by a computer in the first embodiment. FIG. 10 is a functional block diagram illustrating a configuration of a computer according to a second embodiment. 3 is a flowchart illustrating an operation of a process compression process by a computer in the first embodiment. FIG. 10 is a functional block diagram illustrating an example of a configuration of a computer according to a third embodiment. FIG. 10 is a functional block diagram illustrating an example of a configuration of a computer according to a third embodiment.

Explanation of symbols

1 Computer 11 Process Execution Processing Unit 12 Compression Dictionary Selection Processing Unit 13 Compression Processing Unit 14 Compressed Data Storage Processing Unit 15 Process Restoration Processing Unit 21 Processes 22 and 32 Compressed Data Storage Unit 31 Compression Dictionary Storage Unit

Claims (19)

An information processing apparatus that executes a predetermined process on a main memory,
A process compression means for compressing an image of a process being executed on the main memory based on a compression rule corresponding to the process being executed;
An information processing apparatus characterized by that. Compression dictionary storage means for storing compression dictionary data representing the compression rule for each process;
Compression dictionary selection means for selecting the compression dictionary data corresponding to the running process to be compressed,
The process compression means compresses the image of the running process using the selected compression dictionary data.
The information processing apparatus according to claim 1. Compression dictionary generation means for generating compression dictionary data representing the compression rule corresponding to each of the running processes to be compressed,
The process compression means compresses the image of the running process using the generated compression dictionary data.
The information processing apparatus according to claim 1. Comprising compressed data storage means for storing the compressed data compressed by the process compression means in a predetermined storage device;
The information processing apparatus according to claim 1, 2, or 3. The compressed data storage means stores the compressed data in a predetermined auxiliary storage device;
The information processing apparatus according to claim 4. The compressed data storage means stores the compressed data in the main memory device when the compressed data can be compressed to a predetermined ratio or less compared with the image before compression, and otherwise stores the compressed data. Store in a predetermined auxiliary storage device,
The information processing apparatus according to claim 4.   The information processing apparatus according to claim 1, wherein the compression rule is set according to a type of the process being executed.   8. The information processing apparatus according to claim 1, wherein the compression rule is set according to a virtual address set for the process being executed.   The information processing apparatus according to claim 1, 2, 3, 4, 5, 6, 7 or 8, wherein the process is a process having a part of information unique to each process. An information processing apparatus is a process compression method for compressing a predetermined process executed on a main memory,
A process compression step of compressing an image of a process being executed on the main memory based on a compression rule corresponding to the process being executed;
A process information processing apparatus. The information processing apparatus stores in advance compression dictionary data representing the compression rule for each process, and includes a compression dictionary selection step of selecting the compression dictionary data corresponding to the process being executed to be compressed. ,
The process compression step compresses the image of the running process using the selected compression dictionary data.
The process compression method according to claim 10. A compression dictionary generation step of generating compression dictionary data representing the compression rule corresponding to each of the running processes to be compressed;
The process compression step compresses the image of the running process using the generated compression dictionary data.
The process compression method according to claim 10. A compressed data storage step of storing the compressed data compressed in the process compression step in a predetermined storage device;
The process compression method according to claim 10, 11 or 12. The compressed data storing step stores the compressed data in the main memory device when the compressed data can be compressed to a predetermined ratio or less compared with the image before compression, and otherwise the compressed data is stored. Store in a predetermined auxiliary storage device,
The process compression method according to claim 13. To an information processing device that executes a predetermined process on the main memory,
A program for realizing a process compression means for compressing an image of a process being executed on the main memory based on a compression rule corresponding to the process being executed. Compression in which the information processing apparatus stores compression dictionary data representing the compression rule for each process in advance, and the information processing apparatus selects the compression dictionary data corresponding to the running process to be compressed While realizing dictionary selection means,
The process compression means compresses the image of the running process using the selected compression dictionary data.
The program according to claim 15, wherein: The information processing apparatus realizes compression dictionary generation means for generating compression dictionary data representing the compression rule corresponding to each of the running processes to be compressed,
The process compression means compresses the image of the running process using the generated compression dictionary data.
The program according to claim 15, wherein:   18. The program according to claim 15, 16 or 17, wherein the information processing apparatus is caused to realize compressed data storage means for storing the compressed data compressed by the process compression means in a predetermined storage device. The compressed data storage means stores the compressed data in the main memory device when the compressed data can be compressed to a predetermined ratio or less compared with the image before compression, and otherwise stores the compressed data. Store in a predetermined auxiliary storage device,
The program according to claim 18, wherein:

Images