JPH01166247A - Data reference system for decentralized system - Google Patents

Abstract

PURPOSE:To improve search processing efficiency for subject data by searching sites containing data in the order of higher data access frequencies. CONSTITUTION:Plural sites of a decentralized information network system are defined by the personal computers 3 and 8, the work stations 2, 4, 5 and 9, and the servers 6 and 7, etc. When an access is given to the data son said network system, the presence or absence of data is successively checked for those sites via a control part in the order of higher application frequency. The rate of presence is increased for the access subject data at the reference site. Thus it is possible to improve the search processing efficiency for the subject data in a short period and with a small number of processes.

Description

[Detailed Description of the Invention] [Industrial Field of Application] The present invention is applicable to files of multiple sites (workstations, V-bars, etc.) connected to an information network system.
The present invention relates to a method for accessing data on a database or system, and particularly to a data reference method suitable for selecting a site where data to be accessed exists.

[Conventional technology]

Regarding file (data) reference methods in conventional 1 distributed systems, see Information Processing Society of Japan Journal vot 27゜1,
This is discussed in a document entitled "Method for configuring distributed processing O8 by networking file management functions" published in January 1986.

However, when the site salt where the file name exists is unknown,
No consideration was given to efficiently searching the site.

[Problem that the invention seeks to solve]

An object of the present invention is to provide a method for efficiently searching for a site containing reference target data when referring to data at each site on an information network system.

[Means for solving problems]

When accessing data on a distributed network system, a user accesses the data without specifying which site the data resides on. Distributed file systems and database systems must determine which side the target data resides on, and this is done by checking each site's dictionary or adjusting to each site.

In the present invention, since frequently accessed sites change dynamically depending on work, a method is incorporated in which the site is searched from a dictionary of frequently accessed sites and the site is adjusted to frequently accessed sites.

〔Example〕

An embodiment of the present invention will be described below with reference to FIG.

FIG. 1 shows the configuration of a distributed information network system to which the present invention is applied. The two systems in my mind are multiple workstations WS, 4.94' personal computers 3.8 and data shared by their information processing devices connected on network 1.

It stores programs and is configured by connecting a shared high-performance printer. Work is carried out between these devices while interacting with each other while communicating, creating, and displaying various information. When a distributed network system is viewed from a single WS or PC, the entire system is viewed as one WS or PC's environmental resources. In other words, it has the characteristic of allowing access without being aware of what data, files, and databases exist on each site.

FIG. 2 is a diagram showing an explicit hardware configuration of a personal computer or WS to which the present invention is applied. On the bus 10, there is a processing device 11, typically a microprocessor. Memory 1 for storing programs and data, communication control device 1 for communicating data with other sites via the network
3 is connected. In addition, there is a disk 15 that stores and manages databases, files, programs, etc., a disk control device 14 that controls it, a keyboard 16 that inputs commands to give work instructions to a personal computer or WS, and a keyboard 16 that inputs programs and data. A display device 17 for displaying data and processing results is connected.

This bread/or WS accesses files and databases on the disks of computers and WSs (hereinafter abbreviated as computers and WSs are also included) at other sites via the network.

At this time, in one distributed network system, file A
can be used without specifying which site is located, and the program on the pan controller must determine which site the file entered from the program or keyboard is located at. The same applies to database table names and mailboxes that store mail. In this embodiment, the present invention will be explained by taking data reference called file access as an example. FIG. 3 shows the data structure of the file system. File data 35 stored on the disk 15 is identified from programs, commands, etc. by a file name (string of characters) indicated by 32. One file is.

An IJ is prepared in which a pointer 33 to an area storing a file name 32 shown in 31 and attribute information 34 indicating the nature of the file is stored. Attribute information includes file size 36. File format (continuous file, random file, etc.) 37゜ Area that stores permissions to create, reference, and update this file, location 40 that indicates from which address on the disk the file data is stored, etc. It is configured.

FIG. 4 is a diagram showing that file information exists for each site. My site is.

From 41 entries File A: 4, File B:
43. File C:44... exists, site 1:4
5 are files A1:47.5, respectively. File B1:4
8. File C1: consists of 49... This information is stored on the WS and PC on the network shown in Figure 1.
It is stored on the disk 15 on the server and is used for file access processing.

The file management program extracts the file to the memory 12 in FIG. 2 and refers to it. Therefore, when referencing a file name that does not exist on your own site, you must first check which site the file is on. The time it takes to process will vary depending on which night you start your search. Therefore,
The present invention is based on the idea of starting the search from frequently used sites.

FIG. 5 shows an example of accessing a file by a program. The file is.

As shown in sentence 51, preparations for calling it open are made before use. Here, a file name 52 is given, and a variable called a file descriptor, shown as a result 53, that stores file information used in sentences such as 54.58 is given as a parameter.

In this variable, the file management system organizes and stores data obtained from the attribute information 36 in FIG.

The next 54 statements extract 30 bytes of data from the file to B
This is a statement for reading into a storage area in memory called UF. 55 specifies the file using the file descriptor obtained in 51 instead of the file name. When accessing a file at a site other than the local site in a distributed system, the site identifier is also stored in this file descriptor.

The statement No. 58 processes the file when the reference to the file ends, and makes the file use complete. When using it again, 5
1 0PEN processing must be performed. In this way, if you access the file in the program without specifying the site salt, you do not have to change the program even if the site changes.

Figure 6 is a flowchart describing the process of searching each site and determining which site the file belongs to when identifying a file by file name, as in statement 51 in the program shown in Figure 5. .

61 searches for an entry in the file name list shown at 41 in Figure 4, and there is a file name shown in column 32 that matches the file name you are looking for (like the file name at 52 in Figure 5). Find out.

If it exists, the entry of the matching file is returned in 63 as a return value. The returned program performs processing such as constructing descriptors. 62, if it is not on your site, search for another site.

Since the site you are looking for has a list of sites other than your own, the communication control device 13 sends the file name to the other site via the network 1 and asks it to search for files in that site.

In step 65, it is checked whether the site to be searched still exists. A return value of 66 indicates that the remaining applicable file name is undefined on the distributed system. When there is a site to search.

As mentioned above, the process 67 is to search for the corresponding file name. The search method uses the sent file name and compares it with column 32 of the file name list 31 in FIG. 3 to determine if it matches. Step 68 is for determining a response from the site that requested the search to see if the corresponding file exists. 69
is a process when there is a file name that matches the site, and related information such as the file name entry and site salt is returned as a return value. If the file corresponding to the site does not exist in the judgment in step 68, the process in step 65 is to check if there is a site to be searched next in the list of sites within the own site.

FIG. 7 shows a data structure used to search for sites in which a file name is searched, starting with the most frequently referenced sites.

The list of sites that can be accessed on the network system is 75, and 72 is the site salt% 73 is?
- Entry of the next site to be searched (column 771 stores, 74 indicates a column that stores the entry of the site to be searched immediately before the site in question. 73,740 A column value of 0 means that a connection is made before this) It shows that there is no list of sites that are available.

Figure 7 shows columns 73.7 of entry number 71, 4 and 3.
4 is its value. Column 73 of this table 750 serves as a front chain, and column 74 serves as a rear chain, and the user's own search can trace the previous or next site using this chain. 70° en) IJ
indicates the entry of the site located at the top, and the list of site salts in column 73 is constructed from the most recently referenced site via 70. This allows you to search for file names from frequently referenced sites.

FIG. 8 shows a flowchart of the process of searching for a file name using the data structure of the site name list shown in FIG. 81 is.

70 in FIG. 7 shows the process of extracting the first entry of the site name list.

Judgment 82 indicates that when the value of the retrieved entry is 0, there are no more sites to search.
After all, it shows that the file name is not defined on the distributed network system.

Therefore, in the process 83, a return value indicating that the file is undefined is returned.

82, when the value of the entry is not 0.

Indicates that the site to search for files exists.

Therefore, take out the site salt from column 72 of list 75, send the file name to that site, and have it search within that site. the result.

When there is no file name that matches the site in the judgment of 85,
The process at 86 is to extract the value of the entry of the next site to be searched from the column 73 of the corresponding entry at 75.

If the corresponding file exists in the site searched by 85, first check the first site in the list by 87. This can be determined by checking whether the entry value indicated by 70 is equal to the value of the searched site. If they are the same, the search target site salt is returned as a return value in step 99.

If the site is not located at the top of the site list, the site with the file name is moved to the top of the site name list. The processing is the processing from 89 to 98. First, this process of removing the entry of the site from the current position takes out the entry of the previous site from the column 74 of the relevant entry (process 90). Change the forward chain of the entry at the previous site to the forward chain of the entry in question. This is the process 91. Next, the rear chain of one of the front site entries of the site is corrected. For this purpose, in step 92, the entry one position ahead of column 73 is extracted. Column 74 of the entry contains the number of the next entry, so in process 93 the value of column 74 of the entry is stored in column 73 of the entry found in step 92, completing the modification of the backward chain of the previous site. do.

Next, the entry is moved to the top of the site restore 5. 94 is 70 to retrieve the top of the site list
Extract the value of . Then, in step 95, the value of the ninth entry obtained in step 94 is stored in the column 73 of the entry. The front chain is completed.

Next, in order to create a backward chain of the entry found in step 94, the value of the entry is set in column 74 of the entry found in step 94, and the backward chain is completed. At step 97, 0 is set in the column 74 of the entry to mark the end of the rear chain. Further, in step 98, the head of the site list 7 is made the corresponding entry, and the modification of the list is completed. 99 returns the site salt where the file name to be searched is found as a return value, and ends the search process for the site where the file name exists.

〔Effect of the invention〕

According to the present invention, when accessing files, databases, etc. of other sites in a distributed network system,
When it is not known in which site the data exists, the presence of the data is searched from the frequently used sites, which creates the possibility that the data to be accessed is more likely to exist in the referenced site. As a result, it is possible to shorten the response time, reduce the amount of communication in the entire system, and reduce the search processing for target data that occurs at each site.

[Brief explanation of the drawing]

Figure 1 is a block diagram of the system configuration showing the environment in which the present invention is used, and Figure 2 is a block diagram of the system configuration showing the environment in which the present invention is used.
FIG. 3 is a block diagram of the hardware configuration of S, and FIG. 3 is an explanatory diagram showing a file management system. Figure 4 is an explanatory diagram showing the relationship between each site and the file management method, Figure 5 is an explanatory diagram showing an example of a program for accessing a file, and Figure 6 is a site where the file name exists based on the file name. FIG. 7 is an explanatory diagram of a data structure for storing data for realizing the present invention, and FIG. 8 is a processing flow diagram of a process for searching for a file to be searched. Wow! I, yo74.111□tJJ1・'f3 1
Figure '#] 3 Figure v 4 Mouth ■ 5 Figure REAE) (+i, BUF, 30),
-5'CLθSE (chit) ~ yB Fig. 6'fJ7 Fig. 3 Fig. (,A) 15 3 Leap (B)

Claims (1)

[Claims] 1. In computers and workstations equipped with magnetic disks, CRT displays, memory, and arithmetic processing units that access data on computers, workstations, and servers connected to a network, computers, workstations, and servers at other sites 1. A data reference method for a distributed system, characterized in that a control unit searches for sites where data exists in order to access the above data in accordance with the order of sites with high data access frequency.