JP2003271352A - Data management method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that it is difficult to properly predict the future use frequency and an efficient data deletion is not performed under one-sided condition setting since data use conditions are various according to system conditions. <P>SOLUTION: Whether or not a character pattern being an erasure candidate extracted based on a 'next access prediction time' should be erased is judged under the consideration of frequency on a priority judgement reference memory 8. As for character data C whose 'average use time interval' is '6 days' and character data D whose 'average use time interval' is 'one hour', when the 'next access prediction time' of the character data C is a value which is more close to the current time than that of the character data D, the 'mean use interval' is compared with a threshold value for judging whether the frequency of the character data being the erasure candidate is high or low, and when it is smaller than the threshold, it is judged that the frequency of the character data is high. Supporting the 'next access prediction time' is farther from the current time than the other character data, the character data are not set as the erasure candidate but successively registered in a character pattern memory 7. <P>COPYRIGHT: (C)2003,JPO

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for managing data in a computer system, and more particularly to a method for recording data on a storage medium such as a magnetic disk device, an optical disk device, and a memory. Regarding management methods. 2. Description of the Related Art A computer system is equipped with a device for storing and managing data, a memory and the like in all devices constituting the system. For example, data storage devices such as magnetic disk devices and optical disk devices,
Alternatively, a memory or the like for storing data in an output device such as a printer exists. [0003] The storage capacity of these data storage devices has been increasing year by year with the progress of technology. However, with the progress of computerization of all kinds of business in recent years, the amount of data handled by users has increased dramatically. To cope with such an enormous amount of data, it is first conceivable to add a data storage device or a memory, but this is naturally limited. It is also important to effectively utilize a limited-capacity data storage device by managing data and deleting unnecessary data sequentially in addition to the capacity expansion. [0004] However, it is very difficult to extract unnecessary data from a huge amount of data by artificial work.
Accordingly, various methods have been devised so far for computers to determine unnecessary data. [0005] These are inventions relating to output devices such as printer devices and data display devices. In order to print and display characters and data, pattern information of the characters and data is read from a host device or an external storage device, and printing and display are realized. However, since the process of reading the pattern information from the host device or the external storage device is also performed every time printing or display is performed, the time required for the reading has a large effect on the entire processing time, and is extremely inefficient. is there. Therefore, generally, once read pattern information is stored in a memory, and when the pattern is used for the second time or later, the pattern information that has been registered in advance is used instead of being read from the host device or the external storage device. Printing and display. However, as described above, since the storage capacity of the memory is limited, when new pattern information is added in a state where almost all the area of the memory is used and the storage capacity cannot be ensured, the problem of how to perform the registration. Occurs. At this time, the most common method is to extract and delete pattern information that is predicted to be the least frequently used in the future, and to register new pattern information in the area secured thereby. As described above, in order to effectively utilize the storage capacity of a data storage device, the priorities of pattern information are sequentially determined, and pattern information having a lower priority is deleted. In this case, a method of securing a storage capacity for registering new pattern information is adopted. As a criterion for determining the priority, for example, a method of calculating “frequency of use of pattern information in a certain period” from “number of times of use of each pattern information” and “elapsed time since last access time” and setting a low frequency as a deletion target Is generally used. However, the conditions under which pattern information is used vary depending on the form of the system, and in some cases, appropriate judgment cannot be made only by the frequency of use or the elapsed time parameter. [0008] An object of the present invention is to provide a method of deleting unnecessary pattern information by using "the number of times of use of each pattern information" or "elapsed time since last access time" as a method for deleting unnecessary pattern information. An object of the present invention is to provide a data management method capable of appropriately predicting “pattern information that is considered to be the least frequently used in the future” even when it is difficult to make prediction. In view of the above problems, a data management method according to the present invention records last use time information, cumulative use number information and average use interval for each of a plurality of stored data. Regarding the management method of the data storage means for holding and updating these information at the time of using each data or at regular intervals, when registering new data, there is a sufficient free space in the data storage area of the storage device. An empty area determining step of determining whether or not there is an empty area in the empty area determining step,
Additionally registering the new data in the data storage area,
A new registration step of recording the last access time information, the cumulative use count information, and the average use interval information; and, if the free area is insufficient in the free area determination step, the next access prediction time of the plurality of registered data. Extracting the most distant data, a deletion data extraction step as a deletion candidate, and a frequency comparison step of comparing the average use interval of the data deleted as a deletion candidate in the deletion candidate extraction step with a threshold prepared in advance. In the frequency comparison step, when the average use interval of the data set as the deletion candidate is larger than the threshold value, the data set as the deletion candidate is deleted from the data storage area; A deletion data determining step of excluding the data set as the deletion candidate from the deletion candidates; and a deletion data determining step. After completion of the backup, it is determined again whether or not there is a free area, and if the free area is not enough, in the deletion data extraction step, data farthest from the next access predicted time is extracted as a deletion candidate. . Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing an example of a data storage device to which the data management method of the present invention is applied. 1 is a computer, and 2 is a printer control device placed under the control of the computer. The printer control device controls the printing process on behalf of the computer 1 and executes printing from the printer 3 connected to the computer 1. The detailed internal configuration of the printer control device 2 will also be described with reference to FIG. 4 is a printer controller 2
Is a CPU for controlling the internal processing of the printer, and 5 is a ROM in which a print processing program is stored. Reference numeral 6 denotes a character management table memory for managing the use status of character data, 7 a character pattern memory for storing character pattern information, 8
Is a priority determination criterion memory for determining whether or not pattern information that has become a deletion candidate should be deleted from the viewpoint of use frequency. The character management table memory shown in FIG. 2 is a memory for managing the use status of character data on the character management table memory 6. When a print request is issued from the computer 1 to the printer 3, the character management table memory 6 registers the character codes of all character data. At this time, if the same character code has already been registered in the character management table memory 6, "last access time", "average use interval", and "predicted next access time" for this character code are calculated and registered. The "last access time" is the latest use time of character data, and is updated and registered at the processing time (= current time) every time a print request is generated. From FIG. 2, the last access times of the character code A and the character code B are “12:00 on July 1, 2001” and “200”, respectively.
It is found that the time is 13:00 on July 6, 2001. The use time interval of the character code is obtained at the time of update of the last access time or at regular intervals by taking the difference between the last access time registered so far and the current time. An “average use interval” is further calculated and registered based on the cumulative use count of the character code. As a matter of course, when registering a new character code, the “average use interval” cannot be calculated. In this case, “0 (zero)” is temporarily registered for priority determination described later. In the example shown in FIG. 2, the average use intervals of the character codes A and B are “7 days and 1 hour” and “5 hours”, respectively.
Days and 3 hours ". The “predicted next access time” is a value obtained by adding the “average use interval” to the “last access time”. In the example shown in FIG. 2, the predicted next access times of the character code A and the character code B are “13:00 on July 8, 2001” and “16:00 on July 11, 2001”, respectively. The character pattern memory shown in FIG. 2 is a diagram showing an internal area of the character pattern memory 7, and a character code and its character pattern information can be registered in association with each other. The priority criterion memory of FIG. 2 is a diagram showing an internal area of the priority criterion memory 8. In the priority determination criterion memory 8, it is determined whether to delete the character pattern information of the deletion candidate extracted based on the “predicted next access time” in consideration of the frequency of use. For example, regarding the character code C whose “average use time interval” is “7 days” and the character code D of “1 hour”, if the character code C
If the “predicted next access time” is a value closer to the current time than that of the character code D, the character code D is not used as a deletion candidate instead of using only the “predicted next access time” as a deletion candidate. It can be said that making the character code C a deletion candidate is more appropriate as a data management method. A threshold for determining whether the use frequency of the deletion candidate character code is high or low is stored in the priority determination reference memory 8 in advance. In this example, the threshold is temporarily set to “12 hours”. In this case, if the “average use interval” is within 12 hours, even if the “predicted next access time” is farther from the current time than other character codes, it is not registered as a deletion candidate and is continuously registered in the character pattern memory 7. Hold. Next, the operation of the present invention will be described in detail with reference to FIGS. It is assumed that the current time in this embodiment is 10:00 on July 7, 2001. First, the computer 1 sends a character code to be printed to the printer control device 2. Upon receiving the character code, the printer control device 2 checks whether the character code exists in the character management table memory (step 1). When the character code exists in the character management table memory, the character pattern information corresponding to the corresponding character code is read from the character pattern memory, and the printer 3
And print processing is performed (step 2). Further, the following processing is performed on the corresponding character code in the character management table memory. In the “last access time”, the time at which the print request is made this time (current time) is updated and registered. The “average usage interval” is calculated based on the average usage interval value up to the previous time, the usage interval value from the last access time to the current time (current usage interval value), and the cumulative usage count so far. Update and register the value. The calculation formula is shown below. ## EQU1 ## The process of updating the “average use interval” is performed not only when a print request is issued, but also when a new registration is made and every predetermined period (for example, every day at 0:00). First, at the time of new registration, the average use interval up to the previous time is 0, and the current use interval (current time-last access time) is also 0. Since this is divided by the cumulative use count “1”, the “average use interval”
Is 0. In the update every predetermined period, a predetermined value is cumulatively added to the current use interval value. However, since the cumulative use count and the average use interval up to the previous time do not change, the value of the average use interval is always large. Become. When a print request occurs, the cumulative use count is incremented by 1 and then divided, and the obtained average time interval is updated and registered. The calculation formula is shown below. ## EQU2 ## As the “predicted next access time”, a value obtained by adding the “average use interval” recalculated to the current time is updated and registered (step 3 above). Conversely, if the received character code does not exist in the character management table, it is checked whether there is an empty area in the character management table (step 4). Requests transmission of character pattern information of the corresponding character code (step
5). The printer control device 2 receives the character pattern information transmitted by the computer 1 (step 6), sends it to the printer 3 (step 7), and performs a printing process. Thereafter, the “last access time” is set in the free area of the character management table memory (step 8), and the character code and the character pattern information are updated and registered in the free area of the character pattern memory (step 9), and the process is terminated. If there is no free area in the character management table (step 4), the processing shifts to a deletion candidate extraction process. First, on the character management table, search for a character code whose “next predicted access time” is the furthest from the current time (s
step10). In this process, the character management table memory is in a state as shown in FIG. Since the “predicted next access time” of the character code B is furthest from the current time at 16:00 on July 11, 2001, this character code B is selected as a deletion candidate. Subsequently, the "average use interval" of the character code B which is a deletion candidate is compared with a threshold value (12 hours) registered in the priority judgment reference memory (step 11). If the “average use interval” of the deletion candidate character code information is larger than the threshold value, it is determined that this character code is infrequently used, and the deletion is determined (step 12) from the character management table memory 6 and the character pattern memory 7. Delete (s
In step 13), the process proceeds to step 5 to start determination on the next candidate character pattern A. Next, as another example, the case where the character management table memory is in the state shown in FIG.
The processing after step 4 will be described. The character code farthest from the "predicted next access time" in the character management table memory is searched. In this case, the “next predicted access time” of the character code D is “20”.
10:50 on July 7, 2001, which is the farthest from the current time, and is temporarily set as a deletion candidate (step 1).
0). However, the "average use interval" of the character code D is one hour, and is compared with a threshold value (12 hours) registered in the priority determination reference memory (step 11).
Is frequently used and is determined not to be deleted from the memory, and is removed from the deletion candidates again (step 1).
4). Subsequently, a search is made for a character code having the next distant access time in the character management table memory (step 15). In this case, since the “next predicted access time” of the character code C is the next farthest, the character code C is set as a deletion candidate, and the process proceeds to step 11. At this time, the "average use interval" of the character code C
Is six days (144 hours), which is larger than the threshold value registered in the priority determination reference memory, so that the character code C is determined to be infrequently used, and is determined to be deleted. As described above, in the data management method of the present invention, in order to effectively use the capacity of the data storage device and the memory, the judgment for deleting the infrequently unnecessary data at any time is made. Can be performed more accurately. In particular, since the deletion target is determined on the basis of the average use interval of each data, it is more in-house than a system handling information of an unspecified number of people, such as a computer system for mail order sales. Like a business system, data to a certain extent is defined daily, weekly, ...
A more remarkable effect is shown in a system in which the processing is repeated at regular intervals as shown in FIG.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 shows an embodiment of a data management method according to the present invention.
It is a block diagram showing an example of composition of a system to which this method is applied. FIG. 2 is a diagram illustrating a configuration example of a character management table memory, a character pattern memory, and a priority determination reference memory. FIG. 3 is a flowchart showing a process when character pattern information exists in a table memory. FIG. 4 extracts character pattern information that is infrequently used,
It is a flowchart which shows the process which determines whether to delete. FIG. 5 is a diagram illustrating an example in which character pattern information that has become a deletion candidate from a predicted next access time is excluded from deletion candidates at an average use interval. FIG. 6 is a diagram illustrating an example in which character pattern information that is a deletion candidate from a predicted next access time is determined to be deleted based on an average use interval. [Description of Signs] 1 Computer 2 Printer control device 3 Printer 4 CPU 5 Program storage ROM 6 Character management table memory 7 Character pattern memory 8 Priority judgment reference memory

Claims (1)

Claims 1. For each of a plurality of stored data, last use time information, cumulative use number information and average use interval are recorded and held, and the information is used at the time of use of each data or at regular intervals. A method of managing data storage means for updating the data storage area, in registering new data, determining whether there is a sufficient free area in the data storage area of the storage device; and determining the free area. A new registration step of additionally registering the new data in the data storage area and recording the last access time information, the cumulative use count information, and the average use interval information if there is a free area in the step; If there is not enough free space in the step, of the multiple data already registered, A data extraction step of extracting data and selecting deletion candidates as deletion candidates; a frequency comparison step of comparing an average use interval of the data set as deletion candidates in the deletion candidate extraction step with a prepared threshold value; In the frequency comparison step, if the average use interval of the data set as the deletion candidate is larger than the threshold, the data set as the deletion candidate is deleted from the data storage area. A deletion data judging step of excluding the data set as the deletion candidate from the deletion candidates; and judging the presence or absence of an empty area again after the end of the deletion data judging step. A data management method characterized in that at step, data farthest next to the predicted next access time is extracted as a deletion candidate.