RU98116059A - LOAD BALANCING THROUGH THE COMPUTER SERVER PROCESSORS - Google Patents

Claims (13)

1. A computer server connected to the network and containing many processors (20, 30, 40), configured to serve at least one user computer connected to the network, the service includes providing information to the user computer (50), when the processors have access to a storage device (120), which stores data for information blocks, the server computer contains: means (80) for selecting a block contained in each processor from the above-mentioned set of processors and used in response to the request os from the user computer (50) to a specific block of information from the processor, in order to select data from a storage device (120) to generate a file representing the requested block of information, the file contains links to subsequent blocks of information that may be requested by user (50); means (100) for determining the load for periodically determining activity data for processors to include them in a load distribution record stored for all server processors; means (90) for making a decision, which is such that, for each link to a subsequent block of information in the file generated by the block reading means (80), determine, based on the load distribution record, which processor should serve the request from the user computer (50) to this subsequent block of information, and include the address for this processor in the file generated by the block selection tool (80); and delivery means (110) for transferring the file to the user computer (50).  2. The computer server according to claim 1, wherein the load determination means (100) is included in each processor from said plurality of processors in order to determine activity data for each processor.  3. The computer server according to claim 1 or 2, in which each processor from the aforementioned set of processors is equipped with a decision making means (90).  4. The computer server according to any one of the preceding paragraphs, in which if a plurality of subsequent blocks of information will be requested by the user computer (50) when the information block represented by the file is generated, then the decision tool (90) is configured to determine the number of these subsequent blocks of information, requests for which will be served by any particular processor, and the greater the load on the processor, which is determined in accordance with the record of load distribution, the less the number of information blocks, requests for which this processor will serve.  5. The computer server according to any one of the preceding paragraphs, in which one or more of the subsequent information blocks are image files that are requested by the user computer (50) when the information block is generated.  6. A computer server according to any one of the preceding paragraphs, in which the network is the Internet.  7. The computer server according to claim 6, wherein the service system is the World Wide Web, and the information blocks are web pages.  8. The computer server according to claim 7, in which the address for the subsequent information block included in the file by the decision means (90) takes the form of an address bar that includes both an identifier for this information block and a special identifier for the processor, which, as the decision tool (90) has determined, it will serve a request for this block of information.  9. The method of operation of a server computer connected to the network and containing multiple processors (20, 30, 40) configured to serve one or more user computers connected to the network, the service includes supplying the user computer (50) with information blocks, while the processors have access to a storage device (120), which stores data for information blocks, the method includes the steps: (a) in response to a request from a user computer (50) to a specific block of information from the processor, using this process litter to fetch data from the memory (120) in order to construct a file representing the requested block of information, the file including references to subsequent blocks of information that may be requested by the user (50); (b) using the means (100) for determining the load to periodically determine the data on the activity of the processors for inclusion in the record of load balancing supported for all server processors; c) for each link to the subsequent block of information in the file generated in step (a), determining, based on the record of the load distribution, which processor should serve the request from the user computer (50) for this subsequent block of information, and including the address for this processor to the file generated in step (a); (d) transferring the file to a user computer (50).  10. The method of claim 9, wherein each processor from said plurality of processors comprises load determining means (100) configured to perform said step (b).  11. The method according to p. 9 or 10, in which each processor from the aforementioned set of processors contains decision-making means (90) configured to carry out the aforementioned step (c).  12. The method according to any one of paragraphs. 9-11, in which if a plurality of subsequent information blocks will be requested by the user computer (50) when the information block represented by the file is generated, then said step (c) includes determining the number of subsequent information blocks the requests for which will be served by any particular processor, moreover, the greater the load on the processor, which is determined in accordance with the record of the load distribution, the smaller the number of subsequent blocks of information, requests for which this processor will serve.  13. The method according to any one of paragraphs. 9-12, in which the address for the subsequent block of information included in the file in the aforementioned step (c) takes the form of an address bar including an identifier for this block of information and a special identifier for the processor, which in step (c) was defined as the processor which will serve the request for this block of information.