Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2000-01-04
2002-05-21
Alam, Hosain T. (Department: 2172)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000, C707S793000, C709S219000
Reexamination Certificate
active
06393418
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field
The invention of this application relates to a method of sorting a file and an apparatus therefor, which are suitably applied in order to produce a candidate-file list of the files that in the hierarchy storage management system are moved from a memory whose speed is high and whose capacity is small to a memory whose speed is low and whose capacity is large.
2. Description of the Related Art
In a computer, regarding a memory that is used to store therein files of program data, files of data to be processed (such as text data, image data and the like), etc. there has hitherto been a demand for more increase in capacity as the file size increases. Simultaneously, there has hitherto been a demand for more increase in speed (decrease in the access length of time) in order to shorten the processing length of time of the computer.
With the use of only one kind of memories, it is in reality impossible to make such increase both in capacity and in speed go together. For this reason, generally, in computers, a hierarchy storage management (Hierarchy Storage Management) system has been adopted.
In the hierarchy storage management system, there is used a memory that has a hierarchical construction wherein a high-speed and small-capacity memory (buffer memory) and a low-speed and large-capacity memory (main memory, external memory device) are used in combination. And the processing unit of the computer executes a program for realizing the hierarchy storage management (that is also called “an HSM software” by taking the capital letters of the Hierarchy Storage Management). And the unit thereby investigates whether the file that has been requested to be accessed exists within the buffer memory. If this file exists, the computer reads it out from the buffer memory. If it does not exist, the computer transfers it to the buffer memory from the main memory or the external memory and thereafter reads it out from the buffer memory.
As a result of this, since the file for which access request has been made even once is stored in the buffer memory, regarding this file it results that access is made to the buffer memory. The file for which access request has been made once has a general tendency (the localization of the access) that there is a high possibility of the access request being made again in the near future. Therefore, with the file for which access request is highly possibly made being stored in the buffer memory in this way, the frequency of access to the main memory or external memory device is decreased. Thereby, the average access length of time is shortened (namely, the increase both in capacity and in speed go together).
Meanwhile, since the buffer memory is small in capacity, a space area thereof becomes gradually smaller as long as files are only kept transferring from the main memory or the external memory device and finally it becomes impossible for new files to be transferred to the buffer memory. Therefore, the HSM software also needs to have an additional function of replacing a prescribed amount of files of the files already stored in the buffer memory with files that are newly transferred (e.g., after this prescribed amount of files is moved again from the buffer memory to the main memory or the external memory device, new files are transferred to the buffer memory).
In the conventional HSM software, the production of a candidate-file list of the files becoming an object to be replaced with such new transfer files (namely the sorting of the files that are taken up as candidate files) was performed, for example, as follows.
That is, regarding the respective files within the buffer memory, based on:
(11) the number of lapse days that lapse from the last access day of a file (day),
(12) the weight of the number of lapse days,
(13) the file size (kilobyte), and
(14) the weight of the file size, a calculation of the score is performed using the following equation (a).
Score=(11)×(12)+(13)×(14) (a)
It is to be noted that the weights (
12
) and (
14
) are the ones which represent as an integer the ratio between the number of lapse days under the assumption that 1 day=1 kilobyte, for example and the file size. Also, there sometimes occurs a case where without using these weights the score is calculated by only using the lapse days number (
11
) and the file size (
13
).
And, by re-arranging the respective files in the sequential order from the file whose score is larger, a candidate-file list is produced.
After having produced a candidate-file list in this way, when performing actual replacement, the prescribed amount of files that is inserted in the upper sequential order of this candidate-file list is replaced with new transfer files.
By selecting a file as a candidate in the order of the file whose lapse days number from the day of the last access or the file whose file size is large in the above-described way, the one of the files within the buffer memory for which no recent access request is made (i.e., the file which since it is less likely that access request will be made again in the near future does not contribute to shortening the access length of time very much even if kept stored within the buffer memory), the one which occupies a large area (i.e., the file which gives a great effect on the enlargement of a space area by being replaced), etc. are preferentially replaced with new transfer files.
However, even the file whose lapse days number from the day of the last access is small is not always requested to be accessed again in the near future. Namely, among these files there can also exist the one whose last renewal day (the day on which the contents of the file were renewed last) is pretty old and which although the user recently casually accessed it is not destined to being accessed thereafter for a while.
However, in the conventional hierarchy storage management system, it is impossible that such a file is inserted in the upper order of the candidate-file list, unless its file size is extremely large. As a result, although such a file does not contribute to shortening the access length of time even if the file is kept stored within the buffer memory, it did not happen that such the file was preferentially replaced with a new transfer file.
Also, there may be a case where, of the files within the buffer memory, the user thinks he wants to more preferentially replace the file whose file size is large than the file whose lapse days number from the day of the last access is large, with a new transfer file, or may be a case where, conversely, the user thinks he wants to more preferentially replace the file whose lapse days number from the day of the last access is large than the file whose file size is large, with a new transfer file.
However, in the conventional hierarchy storage management system, as in the above-described equation (a), the score is calculated according solely to only the lapse days number from the day of the last access, the file size, and the weights that are primarily determined from the lapse days number and the file size. Therefore, there was no room for the user to flexibly decide which one of the file whose lapse days number from the day of the last access is large and the file whose file size is large he selectively preferentially replaces with a new transfer file.
Also, as in the case of the directories of, for example, an IBM-compatible personal computer, generally, the operating system (OS) of each of personal computers is equipped with the function of keeping files in custody by sorting them. In many cases, the user keeps a plurality of mutually associated files (e.g., the files associated with the same theme or the ones associated with the same customer) under the same directory. Accordingly, the files under the same directory are more likely to be simultaneously accessed when the user performs his search, etc.
Also, generally, the operating system (OS) is also equipped with the function of restoring the files that have been erroneously deleted.
Hikita Jun
Mishima Akifumi
Alam Hosain T.
Alam Shahid
Frommer William S.
Frommer & Lawrence & Haug LLP
Sony Corporation
LandOfFree
Method of and apparatus for sorting a file does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of and apparatus for sorting a file, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of and apparatus for sorting a file will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2845258