SA99200851A - An internet cashing system and a method and an arrangement in such a system - Google Patents

Abstract

The present invention relates to an Internet storage program, to the organization, and to a method for serving requests to recall Internet information files in an Internet storage system. The program is installed as a two-tier storage system. In order to reduce the pressure on the 130th central storage server, an intermediate system 110 that internally connects the 100 local storage servers for the program to the central storage server. The organization communicates with local storage servers in accordance with the agreement used for communications between storage servers. When calling an Internet information file from the central storage server, the marshal uses SQL. So, Central Storage Server 130 is mainly dedicated to answering direct (sql) queries.

Description

Y

A storage program for the Internet and a method and organization for this program 0 Background of the invention: The present invention relates to a storage device for the Internet; And in a manner and organization to work on the information files of the storage program in the Internet.

World) in recent years and its most used manifestation; The World Wide Web co yl has evolved into a massive source of information. Anyone can submit (WWW WideWeb where Ladd can) any information; such as maturity; the pictures; Voice and image on the network for any computer user to easily view it in any part of the world as long as they are subscribed to the Internet. the world. It is currently known that the use of the Internet exceeds all types of traditional telephones and faxes on most international communication lines. The amount of broadcasting and playback is constantly increasing; But it is slow and expensive, and the need to increase this quantity continues. The content of the World Wide Web is increasing until it has become uncountable; It may include, but a relatively small part of these (1984 summer Sa) a huge number of hundreds of 1 terabytes of information accessed in practice. So in order for © a the information belongs to us, we reduce this large user margin and the latency that occurs when accessing information on the Internet; There are currently different storage techniques used to limit the amount of information that must pass through the Internet; And also to limit the distance that the information crosses. In the field of storing the materials of the international network; or Internet information files; There are two methods with © the simplest form of the main (Server); Client-side and server-side storage Client-side storage is used in practically every Internet program today. A program restores information files that were accessed on the Internet. AY browsing the net stored in the computer v and when the user wants to access a specific information file for the second time; The "browser" retrieves it from its store; it does not look for it in the Internet again. Browser (or store; PA ia) In order to assist the neighboring user, a proxy device method can be used for storage on the client side In this way, a store is installed at the AT node, which is the form of international network authorization, to which a number of neighboring users are connected, this node 0 can for example be a server within the company. To (http) accesses the international network server on the Internet; the client sends a request of the type of authorization connection point; or the international network authorization server; and does not send it directly to a server on the global Internet pate. From that, the authorization server sends the request to the international ala file on the global Internet, and stores the response and then returns to reply to the client. If at 0 information for the first time, it is transferred to the Internet and stored in the global network authorization server store. asi Subsequent requests to the same profile can then be made from any client that connects to a local server and does not have to create M8 requests to the international network server on the network leda international to be done

A global Internet. The authorization server store can be used outside the company boundaries. or another organization by applying the above method to the Internet server's storage; Which 1 is accessed directly or indirectly by a number of clients. 40-71 GB, about cle Depending on the size and homogeneity of the storage user group in the server, the Internet traffic generated by the JIB usage group will VIA bytes as the size of the store (Rabi) of users. Because of the steady growth in the information provided by the Internet 96 0 * 0 and the World Wide Web, it is obvious that it is required to increase the size of the store over time 0 to retrieve the access rate to The network, that is, the percentage of the required information files that are converted, this secures an important benefit for performance and exploitation cally from the server storage. In addition to, or more. With the behavior of 96 Vo, the optimum for the Internet, if it increases, the access rate can be increased to Typical end users; this would require a much larger storage volume, which is currently gigabytes, but also require a very large number of end users, which 400-7001 vo currently amounts to several hundred thousand. The reason for this is that The more terminal users;

The higher the probability that an end user has already accessed the requested file; Especially if the users share the same privileges. Installing a mass storage is very easy with a suitable computer and adequate disk capacity. However, it is also required that the store be able to handle all requests received from participating end users. using current technology; It is not possible for a single processor computer to service requests from several hundred thousand end users. Therefore, several systems have been introduced to address this problem. This was indicated by the names of the leading experts in solving this problem. Cisco Systems proposes; End users are connected to a primary router 0 directly programmed to forward all global network requests to a set or "de pd" of private storage devices; or "storage agent". Each storage agent addresses a subset of the origin network servers This brings the number of storage factors to (TY) sufficient to serve approximately 8,000,080 terminal users. Inctomy Corporation proposes the use of a key, or what is called a layer key, for ve | Redirects all requests for international web pages to the "inktomi structured server". It uses a "cluster" of large computers that all share the same disk storage system. The number of stations in this poll amounts to VT which also serves about 905,008,060 common terminal users. But adding multiple computers connected to the same disk storage system adds other redundancies and requires additional arrangements, i.e. some part of each computer's capacity is wasted in 0 processing requests. Ld Network (Guild) You propose a two-tier storage solution. The system has several local stores near the terminal users. These local stores are connected to Palm Centralized storage using the Internet Storage Convention (ICP) when storage loss occurs at the local level. And when the requested file is in the main storage; It will be forwarded to the main store and from Ye forwarded to the terminal user. If the requested file is not in the central store either, the main store will direct a request to the originating server and send the file to the user

° terminal. The main store if handles Internet Storage Agreement (ICP) requests from local stores and contacts the originating server if the storage in the central store is lost. It is dal to accommodate the increase in capacitance; There can be different central stores at the same time ». Each addresses a set of native servers. This means that local repositories are able to answer: © every request to debug the main repository server. Since this agreement is not subject to a standard; It means that local stores must be acquired by Lele from Network Appliances; You are. All of these solutions face the same constraint that the central storage server has to deal with the massive connectivity problem in one way or another. This results in less server capacity utilization; and lead to difficulties in serving hundreds of thousands of users; Which should be required to get 0 high network access. When more servers are added, the systems will become more expensive and more complex. The complexity of the system adds mud to the AL and leads to less efficient exploitation of the relatively expensive Pad servers. General description of the invention: One of the objectives of the invention is to overcome the obstacles that face the techniques known today in files 1 information stored on the Internet and to provide a solution to store information files in an efficient manner. Another objective of the invention is to provide a solution of how to serve user requests for stored information files retrieved by a storage system quickly and at reasonable costs. Another goal is to provide a store server solution capable of coping with the increasing number of information files provided by the Internet and the World Wide Web. 0 Another goal is to provide a solution to obtain a high percentage of Jaa number on calling information files directed to a storage system at the lowest cost. Another objective of the invention is to provide a measurable and standardizable storage system. The above objectives are achieved by means of an Internet storage system and a method for serving Internet information file recall requests in an Internet storage system in accordance with the accompanying safeguards. Yo According to the first aspect of the invention; The invention provides a method for serving Internet information file recall requests with an Internet storage system; This method includes steps to receive a request

. retrieve an Internet information file from the user; on the Internet storage server; then respond to the calling request (ca) gl perform a query of said information file; If the information file is not yet stored on the central server; Then the answer to the aforementioned query; directing a subpoena request for the aforementioned information file; The said call is directed to a feeder if for 0. The said answer did not indicate that the central file server that stores the Internet information files contains the stored information file; And then make a query in the aforementioned central server file about the aforementioned information file from the feeding device to respond to the request to recall the aforementioned file; This is in order to reduce pressure on the aforementioned central file server. According to another aspect of the invention; The invention provides a system for storing Internet information. 0 The system includes at least one local server store and at least one local central file server; Both stores store stored Internet information files; Jal relieves pressure on said central file server and includes a feeder that connects to said local storage server and to said central file server where the feeder includes a primary performance for receiving an Internet information file call request from said local storage server; another tool to extract Vo query results from an initial sequence fetched from said local stored server; And third, a performance of the query on the central file server for the Internet Information File using the query: mentioned taken from the second performance. According to the third aspect of the invention; The invention provides an Internet storage system. The system includes a group of local Internet storage servers; where the local server store is designed to receive Internet file recall requests from users; at least one central file server resides in the repository location egg JS all and stored internet information files; feeder device that connects local storage servers to a central file store; The feeder includes at least one feeder; The feeder includes a device to connect to at least one local storage server according to the existing system between storage servers and devices to retrieve Internet information files from the central file 8 server using “lly” base queries, which relieves pressure on the said central file server.

The present invention is based on the idea of connecting a number of computers to a central file server; Or the stored server og JS all that stores internet information files. While Gay is on the stored server cs Spall, these additional computers are low end computers. Private computers are equipped to take the pressure off the central storage server by performing some of the tasks normally performed by the central storage server itself. In this way, the central storage server is able to serve local storage servers that are connected to the central server or connected to the central server through private computers quickly and at reasonable costs. In fact, most of the uses are made with expensive hardware that forms the central file server and its file store in which the files are stored; Inexpensive dedicated hardware around 0 file servers wastes time. The invention of the feeder device was achieved by devices separate from any Al realizing the central file server. This reduces pressure on the central file server, which is able to devote more processing time to the actual recovery of stored information files. If a centralized file server is able to effectively serve a large number of users. Since the 1 file callback requests made by users are served by local stored servers more efficiently, the number of user callbacks served can increase; This in turn causes the central file server to have a higher by idl hit ratio relative to its store. According to one embodiment of the present invention; The change tool communicates with local storage servers; for a central file server; According to the existing connection agreement between Internet storage servers. Yo The system currently in use is either the Internet Storage Agreement (ICP) or the “Digest” cache, but it can be any other traditional agreement or a future agreement used for the same purpose. and query and/or calls for information files on devices separate from the central file server device, the pressure on the central file server will be significantly reduced Yo when a local storage server receives a call from the user for an information file that was not previously stored on the central server; The local server then starts a query for that file

A

Embodiments of the invention direct the query to a table; or database; located within or connected to the central server. If the said table indicates that the file for which the query is being made is stored on the central file server, the local server will retrieve the file from the feeder; or feeder. It is preferable that this call be made in accordance with the absorption warehouses agreement. But; As in the case of calling from the user to the local server; The call from the local store to the feeder can be 0 according to any of the three layer conventions; for example http requests and in another embodiment of the invention; The query is forwarded from the local server to the feeder. The query includes; For example Internet Storage Agreement query (10); on infofile URL scalar The feeder extracts the query count from the URL on which the query was made of type 0. A prefix of the query received from the infofile; this query number Subsequently used by the feeder to query the file server for files used using the query (complex query language) if the file for which the query is made is located on 1 on the central file server, i.e. if there is only one storage access The file on which the query is forced is transferred from the central server, by means of the feeder, to the coated server. In (ICP) as an answer to a query such as a capacity query on a central file server. The raw numeric and from part of the URL information The query number can optionally be extracted from the address in said query. This part of the address information contains information Specifies 0 the language it uses; the JB server enables the user to separate from the original caller ; For central files from the answer according to this specific information, the query number corresponding to the information file is extracted by using the combination of the decimal number system; 1003 is a combination of the decimal system and in an embodiment in which the local server performs an internal query for the information file; Feeder 1 extracts the query number from the next call to the feeder from the local server. sequence

The raw numeric used to derive the query number is the sequence contained within said callback; For example, the URL of the HTTP call (the query number is used if by the feeder when querying the central file server for the information file; it is preferable to use the SQL query) As a basis for deriving the aforementioned query number. In order to take more pressure off the central file server; Preferably, the feeder contains table storage information for each file that is stored in the central file server. Let the table, for example, be a table with a combination of 1405 tabs, and its location is in memory. And by searching in the aforementioned table; It can be deduced whether the information file 0 on which the query is gal is stored on the central file server or not without the need to perform the query on the server; This will enable a faster response from the feeder to the query from the local server. According to another embodiment of the invention; The Internet storage system also includes an update device; or updated; To convert information files that are stored on a central file server. 1 Adee conditioning involves transferring a copy of the file stored on the local server to the central server. The converted file; As a result of a failure to store it on the central server when the file was queried; A file that was retrieved from its original server by the local server and then stored by the local server. if the central file server or central storage server is hurt; does not itself restore an unsaved file; 0, so it is not a burden of jaf to call a file from the original server because of the storage error when the local storage server is running. instead; When the feeder evaluates a call request from the local storage server for an information file; It concludes that the requested file is not stored on the central file server; The feeder routes the answer to the local server running the query; indicates that the file does not exist; Then instructs Shad to modify the central YO file server. after receiving the answer; which indicates a storage failure; The local storage server retrieves the query file from its original server. And after receiving the order to amend

Ve. central file server The updater retrieves a copy of the file from the local server and transfers it to the central storage server where it is stored. It is preferable that the conversion and storage process take place in a time when the general pressure on the central file server is light, as well as at a time when the local server had taken enough time to restore the file from the original server. If the local server is in the secure area; the updater will call a copy of the 0 : file from its original server; This copy is stored on the central storage server. In this case, it is preferable not to order the updated feeder to start the update or switch process except when the clad number of queries received reaches a certain number for the same information file. These queries originate from local servers located behind the security line. It is preferable that the updater be in devices separate from the devices in which the feeders are located; In addition, it must be separated from 0 from the file server. It is useful. Because file recall requests, for example, from origin servers can take an incalculable amount of time and result in HTTP calls that place an incalculable pressure on the devices that perform the calling process. But; In systems, finding the modifier in the same devices that contain the feeders; They are at the same time a simplified Say separate from the central file server. In embodiments in which the updated and federated System_1 is implemented, in which the local storage server communicates internally with the central file server without the hardware itself being present In the central storage location with the central file server, the separation of these devices from the central file server machine is clear. Some Internet information files are not suitable for storage. Such files are sometimes called dynamic information files; the term "dynamic" comes from Those files which are constantly updated -0 on the original server; examples of such files are stock quote files, weather forecast reports, etc. One of the preferred ways to handle the presence of dynamic files is or on the Cad servers to create A list of known files that are not storable either in the local. In this way, as an answer to a query made by the user, communication within the system is reduced to a minimum. ve

: 111 and according to another embodiment of the present invention; Several central file servers are joined to the central storage location; All information files stored on a file server are associated with their original location names; The addresses are extracted from the query number of the information file being called; The feeder responds to a query about server storage files with an appropriate field. In this pre-measurable solution; Each file system server has its own disk; This will relieve pressure. in addition to; © site: central storage is scalable with third-party file servers thanks to the conventions used in this site that define its standards. A quick; It is preferable that each low end computer connects to a server, and if there are several servers of bial (ald) central files by serial 0, then the medium is a private network. In the latter case mentioned, it is preferable to keep at least part of the network capacity For the connection in question. The network being used can, of course, be part of the Internet, and in an unallocated way as well. The type of connection used between the central file server and the low end computers depends to a large extent on the place where you are low end computers; or feeders and updaters; are they in 01 the same location as the central file server or in a different location?. selected from de gana servers In addition, the central storage location preferably serves the local storage; these The group, in turn, serves its culturally and linguistically homogeneous audience.This would increase the percentage of hits in a central storage rate since it is more likely that the same information files are called up more than once.‎ 0 Using the present invention Any operator of the Internet storage system that calls up information files i Pursuant to the present invention; To provide a fast, cheap and effective way to serve a large number of subscribed customers. These customers prefer to be different Internet service providers; or to be companies connected to the invented central storage location or the invented feeders and/or updates to their own local storage servers; Or connected as clients to the system that houses the entire Yo-invented storage system that it creates in a central storage timer that includes the feeders

ip and updates and its storage servers connected to the local. Of course, the client for the client can be Lad a single user who constitutes a single client of the global network and connects directly to the invented system. in addition to; (Sa) Any large company or ISP may have the option to run the invented system for itself without connecting to this system operated by another party. 0 In addition; Lay that the invented storage system is set out in standard agreements; SICP "SQL 0" local storage servers and mainframe file servers from any manufacturer can be attached to the system as long as these conventions are in place. Within the framework of the present invention, the local Internet storage server should be explained as a link 0 paraphrasing link; preferably a link World Wide Web Delegation Restores 0 repository for "users" or web clients. Connected to the Delegation Connection Items stored on the local Internet storage server or file server in the central storage location must be any of the non-dynamic accessible files. Using the Internet and the type of information it includes. If the number of different types of files and the different names of these files are included in the meaning of the Internet information file used in the present Ve invention, such as binary files; text files; audio and video files; HTTP files (conventional Converting bulky text, HyperText Transfer Protocol, World Wide Web (WWW) files, FTP files (File Transfer Protocol), WWW pages, or World Wide Web materials (WWW) and so on. In addition to files that are accessed via an HTTP or FTP agreement, any file that can be accessed from the Internet according to any three-layer agreement can also be included in the meaning of the expression 'information files'. The agreements that can be used is the WTP Agreement (ASL DU Transfer Protocol (Wireless Transport Protocol) which is used within the WAP standard (Wireless Application Protocol). Considering the fourth aspect of the present invention; it includes The present invention is a readable YO computer, in which one or more computer programs store instructions for one computer

one or more general-purpose computers; The purpose of this inclusion is to enable the said computer or multiple computers to perform the steps described in the claims of AV=Y The above aspects as well as other features in addition to the benefits offered by the present invention will be fully explained and understood in the following detailed description; with . © Refer to the drawings accompanying representative embodiments of this invention. Representative embodiments of the invention will be explained below by reference to the attached drawings, which are shown as follows: Figure 1: shows a schematic embodiment of the Internet storage system according to the present invention; 01 0-0 Figure ?: Shows a diagram of another embodiment of the Internet storage system according to the present invention; Figure “: A flowchart of the operations executed by the nal storage server is shown in the figure ?; Figure 4: A flowchart of the operations carried out by the feeder is shown in Figure ?; Figure 10 shows a flowchart of the operations performed by the updater in Figure ?:ve Figure iT shows an additional embodiment of the Internet storage system according to the present invention. Referring to the diagram represented in Figure 1, an embodiment of the present invention is shown. In Jal 1 a number of Jaa local storage servers are indicated. 001 These ad storage servers are 001 by the Internet; By feeder .011 The feeder number 011 and local storage servers number 001 indicated in Figure 1 are just for example; And embodiment is not at all limited to these numbers. But aside from the nutrient numbers; Each feeder in this embodiment is connected to a single central file server. In Figure 1 Feeder 011 is connected to the central file server AT This central file server has a device (not shown) in which the 0 files are stored information; This server is equipped with a high-end computer; Such as DEC Alpha sl Sun Ultra Sparc ©A.H. Each 01011 Ne, on the other hand, is equipped with a lower end computer; like devices

V¢ Classic PC; It includes a front end device that handles the issue of communication between

AY and central file server 001 Local Storage Servers using the Internet Storage Agreement; 001 connects to Local Storage Servers 0 The feeder is a message-based agreement used to communicate between storage servers over the Internet. (Storage Agreement Internet) about the ICP Internet Information File on query 011, feeder 0 replies. And it is an answer from ed oe store; The query is received from one of the local storage servers - 6.0: 10138) either to access the store ICP The ICP answer indicates the type submitted from the canceler on the ICP file according to the Internet storage agreement; includes a query contains a number 011 This infers the searched URL feeder from the file URL info for the query matches the searched info file using the decimal combination Ve -

MD5 combination tabbed by 11 and using the query number; Table 5 is searched

Random sl RAM) contains 011 RAM located in memory. The feeder includes 111 pall tabbed table storage. The Led table, which is 11 (Access Memory), includes an entry for each query number that matches the Internet information file stored in the file server. Searching in the tabbed table 110 means searching for entries for the central AY query number 1 that matches the query number. If any number matching the query is found in the table, then 071 is stored in the centralized file server die This is an indication that the searched file will indicate successful access to the store. local it indicates that the file is 116 and accordingly if a query number is not found in the table and as a result the answer from the AY being searched is not stored in the central file server _ 0 will indicate a loss failure Or failure to store. ICP type The tool to extract the query number using the mixed decimal system 1005 and the search tool in With an appropriate measurement program included in the feeder 171 The tabbed table is the microprocessor The microprocessor runs the measurement program and this results in Operating query number and results in NY The presence of this measurement program is oriented in the form of NYO as well as the search result in the tabbed table Yo Direct to People skilled in the art of programming

Vo has access to the store; the 001 of the local server 011 if the answer from the feeder indicates the local store will retrieve the information file from the feeder using a large text (WWW) transfer agreement; This agreement is used to access the materials of the World Wide Web. It is transferred to the feeder; This request includes a file of type HTTP Internet. That is, a request of type © .URL:

SQL general queries 011 uses the Ve feeder. When connecting to the central file server, the feeder will retrieve the query number that was previously inferred HTTP after receiving a request to extract the query number. ICP for the corresponding URL query for him. ICP is used to query the URL from the SQL file server routing standard The feeder is then used by the query number in the info file query 111 The central file server (cll) will transfer to the central. And as an answer to the 0 subject of the query to feeder 1100; In turn, it forwards the information file to the server that issued the query for the file. 001 local to not stored; the local server 001 is for the local server 011 if the feeder answer of the original server (not shown in the figure) points to the requested file; The HTTP store will direct a request, after which it will receive the file and transfer a copy of it to the user requesting the file (not shown in the figure). in the feeder. The microprocessor also contains a means to receive requests SQL is done using 111 in addition to the query tool on the file server 001 the local server Controlling the operations carried out by the microprocessor by means of a benchmarking program that is part of the performance: mentioned above . The existence of these measurement programs is well known to people who are skilled in the art of programming. They are also aware of the mentioned agreements.‎ The Internet storage system according to the present invention is shown in the form ?. The AT system differs from the embodiment in that the Internet storage system includes an inductor 1 of that shown in Figure 1 in Figure 1 and 701 feeder servers (YY or updater; connected to the central file server Figure ? (YE) shows the inventive device incorporating an on-line oF updater. Yoo vo local storage as well as a feeder. ?0

In addition to what will come from the explanation below regarding the elements of the figure?; The elements of Figure 1, being opposite elements to the elements of Figure 1, work and interact according to what has been shown above with reference to Figure 1. If the images related to the embodiment shown in Figure Y are explained below, as shown below. 0 Updater 7540 is responsible for modifying the storage medium (not shown in the figure) attached to the central file server 770 with the newly stored information files. As shown for Fig. 1; when local server 0010 receives the non-cache message as an answer (o) from feeder 10; as an answer to an ICP query for the same file, local server 000 directs calla For the file: on the original server (not shown in the figure). The requested file is then received and stored on the local server 0.0010 and after a specified time, and as a result of specifying that no storage has occurred in the answer (ACP) feeder 701 directs instructions For updater 740 by modifying the central file server. Updater 746 receives from feed 701 the URL of the file being queried and the identity of the local server Yeo that day the file was queried. The HTTP request for the file is then forwarded to from the updater to the specified local server After receiving the requested file, the updater will store the file in the central file server (771). YT Random Access Memory (RAM) The means of asking for the info file from the local storage server 000 and the storage medium for the info file Receiver on the central file server 771 is the YU microprocessor with the appropriate 0 benchmarking programs contained in the .7400 updater The implementation of these benchmarking programs is a matter known to those skilled in the art of programming. An example of operations performed by the local storage server 7000 in the embodiment sa gall in Figure 1” will now be shown by returning to the flowchart in Figure . In step Fee the local storage server Yoo receives the request for the Internet information file YO 5) the client by a specific local storage server. But the file call request can also be received from the updater oY Eo works according to the explanation of the figure 0 searches the storage server

The local file will then search in step 00 in its locally stored files for the desired file. If the file is found, it will transfer it to the querying client or to the updater (YE). This process is referred to as step JY. If local storage server 70000 does not find the requested file during the search; i.e. did not store the .0 file; It checks at step 07 whether the request came from the updater. If this is true; A message is returned to the server in a step referred to as 04" to indicate that the requested file is not available. If the conditional step 907 is not true, i.e. if the request came from the client, an ICP query is sent in a step referred to as Veo to Feeder .0011 and in the next batch 707 the central store receives an answer of type ICP from feeder 701 indicating whether 0 the central file server 770 has stored the requested file or not.In step 717 the answer is evaluated ICP If the request indicates a loss of storage, i.e. the requested file is not centrally stored, the local storage server 001 directs an inquiry of type 11177 to feeder 0701 about the file; this process is referred to as step Fed and step Next 10; the local storage server receives the requested file from the feeder. Finally in step FV it forwards the file to 1 the client that requested the file. - The operations performed by feeder 70000 in the embodiment shown in Figure Cha 1 return to the diagram The process flow diagram shown in Figure 4. In step 4080, feeder 701 receives an ICP query related to the Internet information file from any of the storage servers. Local Yoo that is handled by the feeder. Jay 0 The query contains Caley's URL The information on which the query is being made. Feeder 0 extracts from this URL in step 500 a query number using the 2005 mix decimal system, and this query number is used in step 407 when searching for the 1405 tabbed mix table in memory 711 of feeder .701. If not The number is found while looking up the combination table; The feeder in step 4007 sends YO an answer of type 10 indicating that it is not stored behind the local storage server 00010 which is where the ICP query came from in step £08 feeder 701 directs a command to the updater Ee restore

What file is the subject of the query not stored; This is done by passing the URL of the file in question to the updater. in step 405; Feeder 701 adds the query number corresponding to the file in question in the tabbed mix table Y10. This is done as a response to the updater's signal to the feeder that the file in question was transferred from the local server ©Yoo and stored by the central file server YW if This updater process 7400 will show JS by reference to Figure 0 if feeder 701 was found in step 07; The number of the query while searching the combination table YO, in step 40 it will send an answer of type 100 indicating that there is a stored operation in. The local buffer server 7000 from which the ICP query came in step 407 is receiving the feeder

ICP that had previously issued query 001 from the local storage server HTTP a request of type 0 for the info file URL contains an HTTP request (ICP) and similar to the requested 0 query. Step of 0A Feeder 701 retrieves the previously extracted query number corresponding to the file.With this query number, the feeder in step £09 queries CFS 7710 for the requested information file using query 501. In step 701 it receives The feeder is an answer from the information file stored from the central file server ;,? Executed by updater 7400 in the embodiment as ?Returning to the flowchart as in figure 0 0 in step 000 ©, updater 7400 receives a command from feeder 701 indicating that a particular file should be fetched. The file fetch of yall was requested by the local storage server Yoo but the feeder found that the storage server (770 Jad) did not store the file. The command contains the URL of the file as well as the address of the local storage server 7000 that requested the file from the central storage 0. The updater will then in step 04) search for the file requested by the command against Yo a list of files known to be unrecoverable for temporary storage. If the list contains the required file; It will be considered void. If the list does not contain the required file; The matter will be suspended

Temporarily by updater 7400 so that local storage server 70000 has time to restore the file from its original server. And at a suitable time for the central file server 177, that is, at a time when the pressure is relatively light on the central server; The central server sends a message to the updater YE saying that it should execute any pending commands; The reception of this message in the Yeo updater is indicated by step 5607. In the next step 0 01; Execution of the command begins with the updater requesting a copy of the file; which is now retrieved and stored locally from local storage server 700 on which Hal originally requested the file. Then a copy of the file is received from the local storage server in step 4. In step 05 ©; The received copy of the file is forwarded to the central file server .YY to be stored there. In the last step 0070 the updater instructs feeder YY to add a query number corresponding to the file stored in CFS 7710 to the hashed mix table V0 The work of CFS 771 is delicate. It mainly performs two roles; It answers SQL queries from feeders 701 by converting the files for these eb aad 0 and it also stores the new information files in its store; And these files are transferred to the store from the updater 500 . Another representative embodiment of the Internet storage system according to the present invention is now shown with reference to Figure +. In Figure 1, is the system different from that shown in Figure 1? In that the system has more than one central file server; The example here shows a system with 71+ storage servers and three cores. Figure 0 Lad + has two feeds + TV each connected to its own local storage server group 01 Feeders 01 and updaters 640 are arranged together such that the central file servers TY are in the central storage location 1960. By means of an “A+ Ethernet” network organized at the central storage location, the updater 6400 and each feeder 01 are connected to all central file servers SY vo The additional number of central file servers in this embodiment secures storage for more files And even more SQL queries to be answered by centralized file servers in comparison

With the embodiment represented in the form?. Since the system can be measured; any number of nutrients; Updaters or central file servers could theoretically be added to the system. The primary difference between the process shown in Figure 1 and the system shown in Figure 7 is that the feeder

Ce needs to select one server from among a number of central file servers 0 : © to which it should direct the SQL query. Each central file server stores 171 information files under their original site names within a predefined domain. If a central file server is selected according to the site name contained in the URL from the central server; either as part of an ICP query or as part of an HTTP request when one of the central file servers is selected by the feeder; The SQL query with the query number is forwarded to the file server NS “CON . 0 It is understood that the structure and function of the elements shown on return to the drawings will be clearer to those skilled in the field. Although the present invention has been explained in terms of specific representative embodiments; There are different options; Explanations and the like will be clear to those skilled in this field. 1 _ The aim of these embodiments is not if the field of invention is limited; The following claims will clarify the intent.

Claims (1)

SAFEGUARDS 1-1 A method for serving Internet information file recall requests in the Y-internet storage system involves the following steps: “| receive the file recall request from the user at the local Internet storage server; . ¢ As a response to a subpoena; A query is made about the said information file © if the said information file has not been stored on the said local server; 1 As an answer to the response to the repeated query, a call request is made for the aforementioned information file 1; Where the call is directed to the feeder if said response indicates that A the central file server that caches files contains the stored info file 4; Ve directs a query; from the feeding device in response to the said coil recall request; To the mentioned central file server 11 to search for the said information file VY in order to relieve the pressure on the said central file server. 1 "- the method as in claim 1; Cus the aforementioned query is executed by Y the central storage server according to one of the conventions used in communication between > Internet storage servers. 1 *- the method as in the claim Cua oF The convention used is the Internet Storage Convention (ICP) 1 ;- Method as in Cum oF The said convention is Y Cache Digest —o 1 Method as in Claims 1-”; where said query is directed by said local storage server Y to said feeder; and said feeder - answers. 1-1 The method as in claim © includes the step of extracting the query number Y In the feed, this information file corresponds to the subject of the query. YY —V 1 The method as in claim 1 where the query step involves the use of Y The number of the query extracted when querying the central file server for a file the said information. 1 +- method as in claim 7; Where said query returns an X: decimal number sequence associated with the information file; and where sequence is used in the mentioned step Extract the query number. 1 4- The method is as in the protection element cA, where the aforementioned decimal number sequence Y Uniform Resource Locator sl URL and where a number is derived Datu) F mentioned from the URL plus at least one part of the header or ¢ The title of the information field for the said query. -0 1 Method as Claims YO) 4 Where the file recall request is submitted " | a number sequence associated with the information file; and where the number sequence is used by YF The feeding device to extract the query number corresponding to the mentioned information file. -11-1 The method is as in claim 0, where the number sequence is the source identifier The italic Y (URL) and where the mentioned query number is extracted from the mentioned URL in addition - to at least one part of the Jia header of the file recall request. 1-1 the manner as in any of the above claims; Jad is on the move Y Creates a tabbed table with an entry for each Internet information file stored on the file server 4 Central. 1 "1-Method as in Claim VY includes the following steps: - perform a tabbed table search for the said information file; f *-- Clarification in the aforementioned answer to the aforementioned query whether the information file 3 mentioned was found during the search process. el al step Cua del method as in any of the protection elements mentioned 6-1 - Query Structured Query Structured yy When the query is made on the said central file server for the said information file (Language 3). ¢ The Cua del procedure step includes the method as In any of the protections mentioned 1-1 the mentioned query has the following steps: And based on the name of the signatory, the original place or the address 7 of the information file; Selecting a central file server ¢ from a group of central file servers; and where each server of the said group © is organized to store Internet information files with their corresponding query numbers 1 within the predefined domain and 7" - perform query on said selected file server for said info file. VY 2 method as in any of the claims Jai 9176-1 on the following additional Y steps: FY local storage server; restore info file said from its origin server 13 3 The reply to his query indicated that said info file was not stored on the central file server °; 1 stored said info file by requesting a copy of the info file from VY's local storage server and storing said copy on the central file server. 0 A updates or modifies the central file server by calling a copy of file 9 information from the local storage server and stores said copy on the central file server 0. 1-1 arrangement in the Internet storage system; the system includes a storage server At least one local Y and at least one central file server; both servers store Internet information 7 files; this arrangement takes pressure off said central file server 3; it also contains a communication feeder with said local storage server 8 and with the central file server aforementioned, where the aforementioned nutrient includes: 1 A first device to receive the Internet information file recall request from the said local storage server; a AT A is a second device to extract a query number in numerical sequence received from LSS-4 - mentioned; And: 0 a third device to query the central file server for the Internet information file 11 using the query number extracted in the second device. 1: 84- The arrangement is as in the VA protection element, where the first device is equipped to operate according to the third layer Internet agreement, Y. -1-01 arrangement as in claims 178-19; Whereas, the third device is configured to use SQL when querying said Internet Information File 3. YY 1 Arrangement as in any of claims 70-178; where the number sequence is contained in the recall request received from said local storage server. 1 '?- the order as in claim 17; where said query is extracted from said number sequence Y plus the header information field of the denied call 1 received from said local storage server. 1'- the order as in the claim (YY) where said query includes a number : a query that is extracted by applying the mixed decimal system to the sequence << mentioned YF and on part of the mentioned header information field. YE 1 Arrangement as in Claims 70-781, wherein the feeder contains: Y A fourth device to receive the query for the Internet information file from the YF central storage server; and ; A fifth device to provide the local storage server with an answer to the received query. Yo 1 Arrangement as in security element (YE) where the said fourth device and mentioned fifth device are equipped to work according to the agreement used to connect between 7 Internet storage servers. 1 - Arrangement as in security element (YO) where the said agreement Y is the Internet Storage Agreement (ICP). Yo ——YY 1 The order as in any of claims 771-74 where the number sequence is implied in said query received from the local storage server. M YA the order as in the claim (YY) where the aforementioned query extracted Y by the aforementioned second device extracted from the aforementioned digital sequence with one part YF : At least for the information field header of the query received from the local storage server. 1 4- Arrangement as in the Cua YA claim, the aforementioned query includes a number \query” and where the query number is extracted from applying the mixed decimal number to the aforementioned dull Y and to at least one part of the header information field. ©0-1 Arrangement as in claim 4 79-7; where gh aa includes Y table with a complete copy of the list of Internet information files stored on said centralized file server "- 1-1 the order as in claim 07; where the answer to the denied query YX is by The mentioned fifth device is based on the content of the mentioned table. YY 1 Arrangement as in Claims 81-171; where the mentioned arrangement; whose purpose is to reduce pressure on said central file server; includes an updater - communicates with said local storage server and with a server said centralized file; where ¢ said updater includes: ° a device for recalling a copy of the Internet information file stored on the local storage server 1; and Y a device for storing the copy it received on the central file server. YY 1 arrangement as in claim (FY) where said pagers Y is configured to fetch a copy of the info file from its original server, if SF the local storage server containing said info file is behind the security ¢ line. SVE 1 Arrangement as in the sandbox 77 or 77, where the updater is equipped to communicate with said feeder Y to receive a command to call said copy About the information file. -Yo 1 the order as in any of the claims 77-;7; Where the updater includes Y List of known and non-ALE information files for storage; which should call copy 7 for these files. - YY 1 Arrangement as in any of the above-mentioned claims (FOV) where “- Said feeder is equipped with a low end computer and the central file server Y is equipped with a higher end computer. glial where FOF (the order as in any of the claims -77 1 Said X is equipped with a lower end computer and the central file server is equipped with a computer 3 ends up. – “317 1 Arrangement as in one of the claims (FO-FY), where the updater is equipped with a computer Y Low end and central file server equipped with a higher end computer. 1 +©- the order as in the FY claim where the updater and at least one feeder Y are equipped with at least one low end computer. 1 -40*- Internet storage system; Includes: Y is a group of Internet storage servers; Where each local storage server is equipped to receive requests to retrieve Internet information files from the user; ; pda At least one central file is located in the central storage location and storage 8 internet information files; And 1 feeder that internally connects a group of local storage servers with the lila server central VY; Said feeding device includes at least one die which in turn includes A on a machine to communicate with the local storage server according to the agreement used for that connection q between internet storage servers and on lea to retrieve internet information files from 0 > said central file server using database queries; leading to 11 Reducing pressure on the central file server. 1 5- The system as in claim 79; Where said feeding device is located at Y is the central storage location. Yv 1 1;- the system as in claim FA 468; Where each feeding device includes several feeders; Each of the feeders internally connects a subset of the *-local storage server cluster with said central file server. 1 "- Internet Storage System as in Claims 41-749; where Y . The central storage is designed to serve specific Ae gana from local storage servers; This Lead, in turn, serves a culturally and linguistically homogeneous community of users. 1 ;- Internet Storage System as in Claims 47-74 where the agreement used is either Internet Storage Agreement (ICP) or Cache | (Digest and 1;- the Internet storage system as in one of the claims 47-74 where each of the Y feeds includes a table with a copy of a list of information files stored in the location of "- the central storage. 1 0 ~ The Internet Storage System as in Claims 44-74 where the Y central file server contains Internet information files with domain names within the specified domain. Addition de includes an update or modification device Connects said central file server to one of the named local storage Y of the group in order to retrieve a copy of the Internet information ¢ file from its origin server or from one of the local storage servers and from ° In order to store the said copy on the central file server.