FR2861865A1 - Backup process to safeguard data includes sending information to remote field terminals for storage, away from source terminal - Google Patents

Abstract

The backup procedure includes requesting memory space in a set of remote field terminals, and determination (E24) of the list of field terminals available. Data is then cut up into segments, prior to transmission to different field terminals. The procedure for safeguarding digital data held in a terminal includes a first request step (E20) to obtain memory space in a set of remote field terminals, accessible via a communication network. The second step includes determination (E24) of the list of field terminals available, and the third step comprises transmission of data to be safeguarded to the available field terminal. The determination of the list of field terminals is carried out the basis of both the list of field terminals used previously, and the list of those connected to the network. A further stage may be included within the process comprising the cutting up of the data to be safeguarded into segments, prior to transmission. The cut-up segments of data may be duplicated, and sent to different field terminals.

Description

Backing up and restoring digital data

  The present invention generally relates to data backup and recovery.

  The user of a computer wishes to save data for later retrieval.

  For this, it is possible to make backups on external physical media such as magnetic tape, diskette, CD-ROM, USB key for example.

  This assumes that the user manages the pace of his backups and the archiving of his physical media.

  It is also possible to call on a service provider, who provides data backup on demand. The data is routed via a communication network to one or more machines operated by the service provider. The network is either an intranet or the Internet.

  However, data backup in a centralized architecture is all the more expensive and complex to implement that the number of user terminals is large. Indeed, the memory space necessary for storing the data is substantially equal to the sum of the memories of all the user terminals. If this storage space is organized centrally, physical limits are quickly reached.

  In addition, a backup on both physical media and a centralized architecture is not perfectly safe. Indeed, the data can be lost in case of theft or destruction of the physical support or the machine of the provider. In addition, the entire data is stored on the same physical space, it is possible to alter, steal or destroy more easily than if the data are distributed.

  It is an object of the present invention to overcome the disadvantages of the prior art by providing a method and device for backing up digital data that allows the simultaneous backup of data from an unlimited number of user terminals.

  To this end, the invention proposes a method for saving digital data of a terminal, characterized in that it comprises the steps of: - a request for obtaining at least one storage space in a set of remote hosting terminals accessible via a communication network, - determining a list of available hosting terminals, - transmitting the data to be backed up to the available hosting terminals.

  Thanks to the invention, the simultaneous backup of the data of an unlimited number of user terminals is possible. Indeed, the more the number of user terminals believes, the more the system is efficient. The backup is performed in a distributed manner in a peer network.

  The system is reliable, secure and its cost is limited. Indeed, it uses the fact that many computer terminals are permanently connected to a communication network and that these terminals often have a storage memory much higher than that required by a user.

  According to a preferred characteristic, the determination of a list of available hosting terminals is performed according to a list of hosting terminals used during a previous backup of the terminal data and / or a list of the terminal terminals connected to the network. Communication.

  The list of available hosting terminals is determined according to real-time characteristics of the network.

  According to a preferred feature, the determination of a list of available hosting terminals is further performed depending on the membership of the terminal to a predetermined community of terminals.

  Indeed, it is possible to limit the list of hosting terminals to a predetermined list.

  According to a preferred characteristic, the backup method comprises a step of cutting the data to be saved in segments, prior to the transmission step.

  The segments are then transmitted to a plurality of terminals, so that each terminal receives only a portion of the data. The risk of piracy or alteration is thus limited.

  According to a preferred characteristic, the data to be backed up are duplicated before the transmission step and the transmission step comprises the transmission of duplicate data to different hosting terminals.

  Thus, the risk of data loss is limited, since they are reproduced in several copies.

  According to a preferred characteristic, the data to be saved are encrypted prior to the transmission step. This increases the security of the data.

  According to a preferred characteristic, at least a portion of the available hosting terminals are determined by rebound.

  The invention also relates to a method for restoring digital data of a terminal, the data having been previously saved in a set of remote hosting terminals accessible via a communication network, characterized in that it comprises the steps of: terminals hosting the data of the terminal, which are connected to the communication network, - recovery of the terminal data from the connected hosting terminals 30.

  According to a preferred characteristic, if no hosting terminal for one or more segments is connected, the rendering process comprises the steps of: - storage of this information, - sending to the client terminal a notification comprising a list of hosting terminals, after the subsequent connection of one of these terminals.

  According to a preferred feature, the rendering method includes the step of decrypting the recovered data, if they were encrypted during their backup.

  The invention also relates to a method of declaring a terminal in a data backup and restoration network, characterized in that it comprises the step of: - reservation of a backup space in the terminal.

  Thus, each terminal is at the same time user or client and host.

  The size of the backup space is determined according to the overall backup space available in the network constituted by the hosting terminals.

  According to a preferred characteristic, the declaration method comprises the step of: inserting the terminal into a terminal base forming part of the network.

  The invention relates to a device for saving digital data of a terminal, characterized in that it comprises: means for sending a request for obtaining at least one storage space in a set of accessible remote hosting terminals via a communication network, means for determining a list of available hosting terminals, means for transmitting the data to be backed up to the available hosting terminals.

  The architecture can be fully distributed or hybrid, with centralized means and distributed resources.

  The invention relates to a device for restoring digital data of a terminal, the data having been previously saved in a set of remote hosting terminals accessible via a communication network, characterized in that it comprises: means for determining terminals hosting the data of the terminal, which are connected to the communication network, means for retrieving data from the terminal from the connected terminal hosts.

  The invention relates to a device for declaring a terminal in a data backup and restoration network, characterized in that it comprises: means for reserving a backup space in the terminal.

  Backup, restore and declaration devices have advantages similar to those of processes.

  The invention also relates to a communication network comprising a plurality of terminals, characterized in that it comprises a backup device, restitution or declaration as previously presented.

  The invention also relates to a computer program comprising instructions for implementing the previously presented methods.

  Other features and advantages will appear on reading preferred embodiments described with reference to the figures in which: FIG. 1 represents a first embodiment of a device according to the invention, FIG. 2 represents an embodiment of FIG. method of declaration 30 of a client terminal according to the invention, - Figure 3 shows an embodiment of connection method of a client terminal according to the invention, - Figure 4 shows an embodiment of the backup method FIG. 5 shows an embodiment of a data recovery method according to the invention, FIG. 6 represents a second embodiment of the device according to the invention.

  According to a first embodiment of the invention shown in FIG. 1, the device is produced in the form of a hybrid architecture that is both centralized and distributed.

  Client terminals TI, T2,..., Tn, connected via an RC communication network such as an intranet, extranet or the Internet, for example, or a mobile network such as GSM, Edge, GPRS, are considered. UMTS for example. These terminals each have a storage space and are part of the distributed architecture. The number of client terminals is unlimited. A client terminal Tn is both a hosting terminal for storing data from other client terminals, and a user terminal for storage spaces in the other hosting terminals.

  The central part e comprises an AD administrator whose main functions are to manage the subscription of the client terminals and to provide them with a dedicated software and possibly a data encryption key to be backed up.

  The central system also includes a GE manager, which manages a database BD subscriber client terminals. The manager also manages a space allocation table TAn for each subscriber terminal T n. This table contains the list of storage spaces allocated to each subscriber client terminal.

  The central system also includes an SU supervisor that maintains a TC table of client terminals connected on the network. Each terminal is classified according to its time of presence on the network. For example, three categories of terminals exist: the so-called "Gold" terminals, which are connected almost permanently on the network, for example at least 12 hours per day, the terminals called "Medium" which have a regular connection, by 5 example of at least 2 hours a day, the "Rare" terminals, which are connected episodically. This ranking is regularly updated by the supervisor.

  The terminals most present on the network will be the most requested by the backup system.

  The operation of the device according to the invention is detailed below using algorithms.

  In the following, we consider more particularly a client terminal Tn. This terminal will successively declare itself in the data backup and restoration network, connect to this network, request a data backup and a data restoration. To simplify the presentation, the hosting function of the client terminal Tn is not considered here. Likewise, the other terminals are considered only as data hosting terminals. Of course, in reality, each terminal is both a client and a host.

  With reference to FIG. 2, a method of declaring the client terminal Tn according to the invention is represented in the form of an algorithm comprising steps E1 to E8.

  The method is implemented in the client terminal Tn which declares itself in the data backup and restoration network, as well as in the AD administrator and the GE manager.

  The step El is the downloading of a software, called client software, backup and data recovery, from the AD administrator to the client terminal T n. This software manages the dialogue between the client terminal and the different modules of the central system, as detailed in the following.

  In the next step E2, the client terminal T sends a request to the administrator AD to request to be part of the backup network and restitution of data.

  The next step E3 is the sending by the administrator AD to the client terminal Tn of a request to ask him for initial information such as: - type of data to be saved: text, image, video, sound, software, etc., level of importance of the data: normal or high, possible choice of one or more closed communities for the backup (company, family, club, etc.), periodicity of backup, full or differential backup, possible encryption, - conservation or not N last saved versions of the data.

  The next step E4 is the response of the client terminal Tn to the administrator AD.

  The next step E5 is sending a request from the administrator AD to the client terminal Tn to reserve a storage space EM n in the memory of the client terminal Tn. This zone is write-protected and read-protected and is under the management of the administrator and the manager. The size of this zone is determined according to the subscription characteristics of the client terminal (step E4) and the storage space currently available in the global backup network. The size of this area is greater than a minimum value that may also vary depending on the overall space available.

  The AD administrator also sends a unique identifier to the terminal Tn.

  The next step E6 is the response of the client terminal Tn to the administrator AD 30. The client terminal can accept or refuse to reserve the space requested by the administrator. It is assumed here that the client terminal accepts.

  The next step E7 is the sending of a request from the AD administrator to the manager GE to update the database TA subscriber client terminals, with the characteristics of the client terminal obtained in step E4.

  In the next step E8, the manager GE updates the database TA customer subscribers subscribers including the client terminal Tn.

  The client terminal T is now part of the data backup and restoration network. It is also said that he is a subscriber.

  With reference to FIG. 3, a connection method of the client terminal Tn according to the invention is represented in the form of an algorithm comprising steps El0 to E14.

  The method is implemented in the client terminal Tn which connects in the data backup and restoration network, as well as in the SU supervisor and the GE manager.

  In step E10, the client terminal Tn sends a connection message to the supervisor SU, by means of the client software.

  In the next step E11, the supervisor SU sends a request to the manager GE to verify that the client terminal Tn is actually authorized to connect to the network. For this, the GE manager searches in step E12 if the client terminal Tn is identified in the base TA subscribed client terminals. It is assumed here that the answer is positive. The manager transmits the response to the supervisor.

  The next step E13 is the reception by the supervisor SU of the answer 25 of the manager GE.

  The next step E14 is updating the TC base of the client terminals connected to the network. The update includes an indication that the client terminal Tn is connected to the network. It is performed by the supervisor SU.

  The client terminal Tn is now in the list of connected terminals which are periodically monitored by the supervisor SU. When the supervisor detects a disconnection, it updates the TC base of the client terminals connected to the network.

  It is now assumed that the client terminal T is connected to the backup network and restitution of data.

  With reference to FIG. 4, a method for saving data of the client terminal according to the invention is represented in the form of an algorithm comprising steps E20 to E28.

  The method is implemented in the client terminal Tn connected to the data backup and restoration network, as well as in the manager GE and the supervisor SU.

  In step E20, the client terminal Tn sends a request to the manager GE to obtain a storage space. This request is sent at the initiative of the client software. The size of the required space depends in particular on the choice of a total or incremental backup.

  In the next step E21, the manager GE establishes a list of terminal hosts, or nodes, used during a previous backup for the client terminal Tn, if there is such a previous backup.

  The next step E22 is the transmission of this list by the manager GE to the supervisor SU.

  The next step E23 is the verification by the supervisor SU of the presence on the network of the nodes of this list. The supervisor SU establishes a list of available nodes, according to the list transmitted by the manager, the nodes present on the network and possibly the membership of the client terminal to one or more community (s).

  In the next step E24, the SU supervisor transmits the list of available nodes to the GE manager.

  The latter updates its allocation tables in the next step E25, according to the list of available nodes transmitted by the supervisor.

  If a previously used host terminal is no longer present, the segments stored on this terminal are marked "free" in the allocation table. This can be done after a certain delay depending on the subscription characteristics, for example concerning the management of the last N backup versions.

  In the next step E26, the handler transmits to the client terminal its allocation table TA ,,.

  The client software then performs, at step E27, a cutting of the files to be saved in segments of configurable size and numbered sequentially. Preferably, the size of a backup segment is smaller than that of a file. Thus, no complete file is transmitted to a remote host, which contributes to the security of the data.

  The next step E28 is the transfer of segments to their respective host. The transfer is carried out in a manner known per se. It is performed according to the allocation table T. For example, segments are transmitted to the terminals TI and T2 and are stored in their respective storage spaces EM1 and EM2. An encryption is possibly carried out before the transfer.

  One segment is duplicated temporarily and the redundant segments are sent to multiple different hosts, depending on the allocation table. The number of copies of each segment is determined according to the level of importance determined by the user (step E4). "Gold" type hosting terminals are more in demand than others. In addition, it is possible to focus on sending redundant segments to terminals located in widely dispersed locations (cities, regions, countries, continents).

  Each host acknowledges the reception and memorization of the segments that were intended for him.

  When the backup of one or more segments fails, for example if the host is disconnected, the client software returns to step E20, with a specific flag recovery. The GE manager then provides the reference of other hosts for the failed segments. The allocation table TAn is updated.

  It is still assumed that the client terminal T is connected to the network for backing up and restoring data.

  With reference to FIG. 5, a data recovery method of the client terminal T according to the invention is represented in the form of an algorithm comprising steps E30 to E36.

  The method is implemented in the client terminal T connected to the data backup and restoration network, as well as in the manager GE and the supervisor SU.

  In step E30, the client terminal Tn sends a request to the manager GE to retrieve the allocation table TAn. The client terminal may optionally send a request to the AD administrator to obtain the decryption key (not shown).

  In the next step E31, the manager GE responds to the client terminal Tn. Similarly, the administrator responds to the client terminal (not shown).

  The next step E32 is the reception by the client terminal T of this information (TAn allocation table and possibly decryption key).

  In the next step E33, the client terminal Tn sends a request to the supervisor SU to retrieve the TC list of the connected hosting terminals.

  In the next step E34, the supervisor SU responds to the client terminal Tn.

  The next step E35 is the reception by the client terminal of this list.

  It should be noted that several hosting terminals for the same set of segments can be connected. They are all registered in the list sent to the client terminal. The "gold" terminals will be solicited first.

  Conversely, if no host terminal for one or more segments is connected, the supervisor stores this information. After the subsequent connection of one of these terminals, the supervisor sends the client terminal a notification comprising a list of hosting terminals so that the client terminal can recover the segments in question.

  In the same situation, it is also possible to provide a temporary storage of these segments, for example in a "gold" terminal. Thus, the management of the presence desynchronization between the terminals is simplified.

  The next step E36 is the recovery by the client terminal Tn of the saved data. Data recovery can be partial or complete. The segments are retrieved and reordered by the client terminal.

  According to a second embodiment of the invention shown in FIG. 6, the device is made in the form of a totally distributed architecture.

  The administrator modules AD ', manager GE' and supervisor SU 'are duplicated on all client terminals T'n. These modules, as well as the client software, are obtained by the client terminal by downloading from a central site, via the communication network.

  The operation of the system is similar to that described above, with some adaptations.

  The supervisor SU 'is active only during the connection time of the client terminal to the network.

  At initialization, the supervisor SU 'has a list of hosting terminals of "Gold" type. Each of these terminals provides a list of connected hosting terminals. The user can possibly modify these lists according to his community of belonging.

  The supervisor SU 'then builds a visibility area that includes a few dozen hosting terminals.

  For the backup, if there are not enough resources available in the visibility space, the supervisor requests the hosting terminals of this space to provide the references of the hosting terminals of their own visibility space. Thus, some hosting terminals are determined by rebound.

  The terminal supervisors work in cooperation.

  The manager GE 'saves the host segment / terminal allocation table with each backup. This recording is stored either on a local physical support, or on a hosting site of the Internet, or according to a backup mechanism according to the present invention.

  2861865 14 When not stored locally, the client software uses the unique identifier to retrieve the host segmentlterminal allocation table.

Claims (19)

CLAIMS,   1. A method for saving digital data of a terminal, characterized in that it comprises the steps of: - request (E20) to obtain at least one storage space in a set of remote hosting terminals accessible via a communication network , - determining (E24) a list of available hosting terminals, - transmitting (E28) the data to be backed up to the available hosting terminals.   2. Backup method according to claim 1, characterized in that the determination (E24) of a list of available hosting terminals is performed according to a list of hosting terminals used during a previous backup of the terminal data and / or a list of hosting terminals connected to the communication network.   3. Backup method according to claim 2, characterized in that the determination (E24) of a list of available hosting terminals is further performed according to the membership of the terminal to a predetermined community of terminals.   4. Backup method according to any one of claims 1 to 3, characterized in that it comprises a step (E27) of cutting the data to be saved in segments, prior to the transmission step.   5. Backup method according to any one of claims 1 to 4, characterized in that the data to be backed up are duplicated prior to the transmission step and in that the transmission step comprises the transmission of duplicate data to different hosting terminals.   6. Backup method according to any one of claims 1 5 to 5, characterized in that the data to be backed up are encrypted prior to the transmission step.   7. Backup method according to any one of claims 1 to 6, characterized in that at least a portion of the available hosting terminals 10 are determined by rebound.   8. Method for restoring digital data of a terminal, the data having been previously saved in a set of remote hosting terminals accessible via a communication network, characterized in that it comprises the steps of: - determination (E34) terminals hosting the data of the terminal, which are connected to the communication network, - recovery (E36) of the terminal data from the connected hosting terminals.   9. A method of restoring digital data of a terminal according to claim 8, characterized in that, if no hosting terminal for one or more segments is connected, it comprises the steps of: storage of this information, - sending to client terminal of a notification comprising a list of hosting terminals, after the subsequent connection of one of these terminals.   10. The method of restitution according to claim 8 or 9, characterized in that it comprises the step of decrypting the recovered data, if they had been encrypted during their backup.   11. A method of declaring a terminal in a backup network and restitution of data, characterized in that it comprises the step of: reserving (E5) a backup space in the terminal.   12. The method of declaration according to claim 11, characterized in that it comprises the step of: - insertion (E8) of the terminal in a base of terminals forming part of the network.   13. Device for saving digital data of a terminal, characterized in that it comprises: - means (Ti) for sending a request to obtain at least one storage space in a set of accessible remote hosting terminals via a communication network, - means (SU) for determining a list of available hosting terminals, - means (Ta) for transmitting the data to be backed up to the available hosting terminals.   14. Device for restoring digital data of a terminal, the data having been previously saved in a set of remote hosting terminals accessible via a communication network, characterized in that it comprises: means (SU) for determining the terminals hosting the data of the terminal, which are connected to the communication network, - means (Ta) for retrieving data from the terminal from the connected hosting terminals.   15. Device for declaring a terminal in a backup network and restitution of data, characterized in that it comprises: - means (AD) for reserving a backup space in the terminal.   16. Communication network comprising a plurality of terminals, characterized in that it comprises a backup device according to claim 13.   17. Communication network comprising a plurality of terminals, characterized in that it comprises a reproduction device according to claim 14.   18. Communication network comprising a plurality of terminals, characterized in that it comprises a reporting device according to claim 15.   19. A program comprising instructions for carrying out the method of any one of claims 1 to 12.