KR0165511B1 - Information test method using sequential link list - Google Patents

Abstract

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information access method in a computer having multiple active processors, and more particularly, to a fast information retrieval method using a hashing function and a sequential link list for accessing information at a high speed, and a fast information retrieval method using a redundant data structure. It is about.

In a computer system with two or more active processors, the same data setting process saves each processor the same information as the data in the exclusive data storage area. The data in the redundant data storage area can be accessed at a high speed by using the coherent hashing function and the link list together to maintain the coherence of the data simultaneously. It is done.

Description

Fast information access method using redundant data structure, hashing function and sequential link list

1 is a block diagram of a device to which a fast information access method using a redundant data structure according to the present invention is applied.

2 is a flowchart of a fast information access method using a redundant data structure according to the present invention.

3 is a flowchart of a fast information access method using a hashing function and a sequential link list according to the present invention.

4 is a data structure diagram of data stored in a redundant data storage area according to the present invention.

5 is a node data structure diagram used in FIG.

6 is a diagram of port information and hashing table structure used in FIG.

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an information access method in a computer having multiple active processors, and more particularly, to a fast information retrieval method using a redundant data structure for accessing information at high speed, a hashing function, and a sequential link list.

In the related art, in a computer-related system in which several active processors exist, the coprocessor must access the information managed exclusively by the main processor.

In other words, to retrieve information managed by the main processor, the coprocessor first asks the main processor for help. In this case, the main processor must stop its work and process the work.

Eventually, this operation interferes with the main processor's main task and causes the main processor to degrade. In addition, since the coprocessor must receive help from the main processor for access to information, there is a problem in that the speed of information access is reduced.

In addition, in the related art, in the case of information retrieval in a data structure connected by a linked list, there is a problem that all lists must be searched in the worst case in order to retrieve necessary information.

Moreover, when such linked lists are sequentially linked by special information in the data structure, there is a problem that it takes a long time to insert or remove one information at a specified position in the linked list.

Although a hashing function can be used to solve this problem, it is easy to insert and remove information using a hashing function, but it still takes a long time to search for information in the next order of the designated information.

Accordingly, an object of the present invention is to improve the performance of each processor by using a redundant data structure that stores information managed by the main processor and redundant information in the information storage area managed by the coprocessor to solve the above problems. It is to provide a fast information access method by using a hashing function that accesses the information at high speed in the data structure and a link list which accesses the information sequentially.

In order to achieve the first object, the fast information access method using the redundant data structure according to the present invention is the same that stores the same information as the data in the exclusive data storage area in each processor in a computer system having two or more active processors. It is characterized in that the data setting process includes maintaining the consistency of data, which simultaneously changes the data in the information storage area installed in a separate coprocessor if the data in the exclusive data storage area is changed.

In order to achieve the second object, a fast information access method using a hashing function and a sequential link list according to the present invention provides a process of searching for a related slot of a hashing table of a node to be accessed, a process of searching a node connected to the found slot, and an access. It includes a fast search process by the pointer of the link list of the node.

Hereinafter, exemplary embodiments will be described in detail with reference to the accompanying drawings.

1 is a block diagram of an apparatus to which a fast information access method using a redundant data structure according to the present invention is applied.

2 is a flowchart of a fast information access method using a redundant data structure according to the present invention. The process includes setting the same data in the redundant data structure of each processor (21) and maintaining the consistency of the set mutual data (22). do.

3 is a flowchart of a fast information access method using a hashing function and a sequential link list according to the present invention. It consists of a search process 32, a search process 33 by a sequential link list.

First, a fast information access method using a redundant data structure will be described with reference to FIGS. 1 and 2.

In FIG. 2, the same data setting process 21 is performed for setting the same data as the exclusive data storage area of the main processor in the redundant data storage area of each active processor.

If the contents of the exclusive data storage area of the main processor are changed after completing the setting process of the redundant data storage area, the contents of the data of the redundant data storage area set for each active processor are simultaneously linked to change the data. It maintains the consistency of the data to maintain sex.

As described above, in the block diagram of FIG. 1, when the coprocessor retrieves the information in the exclusive data area managed by the main processor, the main processor performance is directly retrieved from the redundant data storage area without interrupting the main processor's main operation. In addition to preventing degradation of the system, it is possible to search information without the help of the main processor to speed up information access.

In the block diagram of FIG. 2, in order to maintain the consistency of the data 22, it is necessary to search for, delete, and change the data in the redundant data storage area. .

Next, a fast information access method using a hashing function will be described with reference to FIG. 3.

This is a way to find a node in a hashing table using a hashing function to access a node that contains data, and nodes before or after that node can be found by pointer using the node's link field. .

The above hashing function method is a technique for defining a function that converts a key identifying a data item to information representing a position in a table, and using it to efficiently store the data item in a table or search for a stored data item. Say.

In the first step, a slot search process 31, which is a process of searching for a slot, which is one of hashing tables to which the node is connected, is performed through a hashing function using a MAC address, which is one of information of a node to be found.

In a second step, a node search process 32 is performed to find a node having a given MAC address among nodes connected to a slot which is one of the hashing tables.

After finding the node to be searched in the third step, if the link list before or after is to find the data sequentially, the hashing function does not find a separate node, but instead of searching for a separate node, Search process 33 is performed.

When searching the information of the sequential link list as in the above method, only the first node of the information to be searched is searched by the hashing function, and the rest of the subsequent information is omitted by searching the node by the hashing function and immediately by the value of the pointer. By searching, it is possible to shorten the access time of the material to find. In other words, there is an effect that can access the data you want to find at high speed.

The following describes a specific embodiment of the fast information access method using the redundant structure and hashing function and the sequential link list of the present invention according to FIG. 2 and FIG.

4 shows the structure of data stored in the redundant data storage area mentioned above.

Here, the hash table is a diagram of the hashing table where the hashing function is used, and the free node duplicates the duplicated information when new data is added or deleted in the exclusive data storage area of the main processor. Shows the unit data structure used to store an area. One node data structure is allocated at the free node when needed or returned to the free node when no longer needed. The mac address (mac addr) points to the head node of the node list connected by the sequential link list.

For example, if a node C is to be searched, the slot i of the hash table is first found by using a hashing function, and the node C is searched by using a hash next point of B pointed by I. In this case, if you want to find the C node through the link list of the node, you need to find it using the mac fore field from the node indicated by the mac addr. This indicates that in the worst case, N searches are needed if the number of connected nodes is N.

In the previous example, if you want to find the sequential front and back nodes of a C node, you can simply find the C node and then find the front and back nodes using the mac fore and mac back pointers. In this case, if only hashing function is used, there is a drawback that it cannot find sequential nodes of C node.

Therefore, the present invention solves these shortcomings and improves the speed of access to each data structure.

FIG. 5 is a diagram illustrating one node data structure used in FIG. 4, and FIG. 6 shows a port and hashing table structure used in FIG.

The port information data structure of FIG. 6 is a data structure having information about each port, and a hashing table represents a table structure mapped to a result of the hashing function.

Claims (3)

A method of accessing information of a coprocessor to data in a proprietary data storage area of a main processor in a computer having multiple active processors, The same data setting process for storing the same information as the data in the exclusive data storage area in each processor in the computer system in which the active processors exist; And maintaining the consistency of the data which simultaneously changes the data in the redundant data storage area installed in each processor when the contents of the data in the exclusive data storage area are changed. Fast information access method using a gem data structure, characterized in that the access is performed at high speed. A information retrieval of a data structure connected by a link list in a computer having several active processors, the information retrieval method comprising: retrieving an associated slot of a hashing table of a node to find; A node searching step of finding a node connected to the slot found in the slot searching step; Searching by the pointer of the sequential link list that is searched by the pointer directly from the current node without finding a separate node by the hashing function when searching the data of the link list before or after the node to be searched sequentially. In the case of searching the information of the sequential link list, including only the first node of the information to be searched by the hashing function and the rest of the subsequent information is omitted by the value of the pointer, while omitting the search of the node by the hashing function. Fast information access method using a hashing function and a sequential link list, characterized in that the search can be performed.