Electrical computers and digital processing systems: multicomput – Distributed data processing
Reexamination Certificate
1997-11-18
2002-06-25
Geckil, Mehmet B. (Department: 2152)
Electrical computers and digital processing systems: multicomput
Distributed data processing
C709S245000, C709S246000, C707S793000
Reexamination Certificate
active
06411985
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to a data association apparatus for transmitting data among a plurality of servers, and, more particularly, to a data association apparatus for transferring information about an operation on a database, which has been executed by one transmission server, to a plurality of reception servers.
2. Description of the Related Art
Conventionally, a plurality of servers perform mutually associated data processing on a plurality of databases, such as backup processing, information system processing and batch processing.
The backup processing is executed, for example, for an on-line trading system in a bank. That is, in systems whose stopping services would cause significant social influences, such as a bank on-line system, a backup computer center (backup center) may be built at site remote from the main computer center, in case a computer center (main center) having a regular database for everyday use is downed due to a natural disaster, terrorism or the like. The backup computer center is provided with a backup database having the same contents as the regular database. In the system, every time the server in the main center updates the regular database, the updated contents are transmitted to the server in the backup center. The server in the backup center (which has the same capability as the server in the main center) renews the backup database in accordance with the transmitted updated contents to make the contents of the backup database consistent with those of the regular database. As the consistency of the contents of both databases is maintained, if the main center is downed, the backup center can take over the processing that has been performed by the main center.
The information system processing is executed to collectively manage the results of operations executed in the main business works. The results of operations executed in the main business works are important information for the company activities. Normally, a company builds such information into an information system and positions it as the company's strategic information system (SIS). In this respect, a database for information system is prepared in addition to the databases for the main business works. When server manipulates data for a main business work, the manipulation contents are collectively transmitted to the other server after the work is completed. The other server updates the database for information system by batch processing in accordance with the received manipulation contents. Since the results of a process executed by a main business work concentrate on the database for information system, the results can be managed collectively in the database for information system.
The batch processing is a process which is executed based on the results of the operations that have been carried out by the main business works. In a banking system, for instance, a telephone contact may be made to customers who had made large deposits in deposit and exchange business in the same day, or direct mails may be sent to such customers. Such works are premised on the process of preparing a list of customers who have made more than a certain deposit based on the results of the depositing business in the same day. This list making is executed as a batch process that is performed at the end of the business works in the day. When data of the main business works is manipulated by one server, therefore, the contents of this manipulation are collectively sent to the other server after all the business works are completed. The other server carries out the batch process in accordance with the received manipulation contents.
The procedures for associating data among a plurality of servers in this manner are classified into the following four groups depending on the system in which those servers are built and on the units of data exchanged among the servers.
First, a plurality of servers may be built up on the same computer system and all the records of data are transmitted collectively. In this case, a file which cannot be shared by servers is used in data transmission. More specifically, a transmission server opens such a file, writes all the records of data to be sent into the file, and temporarily closes the file after the writing of the data is completed. Next, a reception server opens this file and reads the stored data.
Second, a plurality of servers may be built up on different computer systems and all the records of data are transmitted collectively. In this case too, a file which cannot be shared by servers is used in data transmission. That is, a transmission server opens such a file, writes all the records of data to be sent into the file, and temporarily closes the file after the writing of the data is completed. This file is transmitted to a reception server by communications means. Upon reception of this file, the reception server opens the file and reads the stored data.
Third, a plurality of servers may be built up on the same computer system and data are transmitted record by record. In this case, a file or memory which can be shared by servers is used in data transmission. More specifically, a transmission server locks such a file or memory, writes one record of data to be transmitted in the file or memory, and then unlocks the file or memory. Next, a reception server locks this file or memory, reads one record of data stored therein, and then unlocks the file or memory. This data writing and reading are repeated until all the records are transmitted.
Fourth, a plurality of servers may be built up on different computer systems and data are transmitted record by record. In this case, a transmission server transmits data to be sent, record by record. A reception server receives the transmitted data record by record. This data writing and reading are repeated until all the records are transmitted.
In any case, the transmission server can send data only to one reception server. That is, the transmission server sends data only once. The transmission server should therefore make this single data transmission at the timing that the reception server could receive the data properly. By the way, there are multifarious types of data requested by reception servers. For example, one server may request the prompt transmission of data while another server may allow for a certain delay and may thus request one data transmission per day in order to suppress the number of processes. Accordingly, the transmission server cannot match the transmission timing with the timings requested by all the reception servers, and can synchronize the transmission timing only with the reception timing of only one reception server. The transmission server cannot therefore send data to a plurality of reception servers.
If a transmission server can output data multiple times simultaneously or in a time-shared manner, the data can be theoretically transmitted to a plurality of reception servers. This system however requires that every time a reception server is changed, the programs or system definitions in the transmission server should be altered. What is more, the burden on the transmission server is inevitably increased. Since the conventional data associating procedures are based on a fixed one-to-one association as explained above, the data flow is not smooth and cannot meet the multifarious requests by a plurality of reception servers. This is the first problem of the prior art.
The first and second procedures described above do not allow the reception server to be operated until the writing of all the records is completed, and cannot ensure parallel running of both transmission and reception servers. The first and second procedures are not therefore designed for the process that is requested immediately. By way of contrast, the third and fourth procedures described above can permit the parallel operation of both servers, but the reception server should be running while the transmission server is sending data. In other words, the transmitted data w
Fujita Kazuhiko
Hashizume Yasuhiko
Kakisu Toshiaki
Shinomiya Yuki
Fujitsu Limited
Geckil Mehmet B.
LandOfFree
Interserver data association apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Interserver data association apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Interserver data association apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2951045