WO2002071222A1

WO2002071222A1 - Method for shutting down a plurality of computers during a power failure

Info

Publication number: WO2002071222A1
Application number: PCT/EP2002/002325
Authority: WO
Inventors: Jens Rehsack
Original assignee: Jens Rehsack
Priority date: 2001-03-05
Filing date: 2002-03-04
Publication date: 2002-09-12
Also published as: DE10110472C2; DE10110472A1

Abstract

A method for controlling a plurality of computers during a power failure, wherein the plurality of computers are connected to each other by at least one network; a first controlling computer is determined from said plurality of computers and performs a predetermined control process when a power failure occurs: an order is established for the plurality of computers in order to ensure that the control process is carried out in an ordered manner; the first controlling computer detects the occurrence of a power failure and performs the control process via the at least one network with regard to the plurality of computers in the predetermined order.

Description

METHOD FOR SHUTDING DOWN A VIL NUMBER OF COMPUTERS WHEN APPEARING

ONE OF ROME'S FAILURE

The present invention relates to a method for shutting down a plurality of computers in the event of a power failure in accordance with the preamble of patent claim 1 and to a method for controlling a large number of computers in the event of a power failure in accordance with the preamble of patent claim 18.

Networked IT systems in which a large number of computers are connected to one another via a network are widely used in industry. In the auto industry, for example, a large number of computers in spare parts stores at car dealers can be connected to a central computer in a central warehouse via a network. Via online connections between the large number of computers and the central computer, e.g. B. Online orders can be executed. In the event that a power failure occurs in the central warehouse and current online orders are not saved in a controlled manner, data loss can occur, which can lead to the IT-supported warehouse systems having incorrect stocks.

Such data losses occur when power failures occur when the computers connected in the networked EDP system are shut down in an unordered manner, i. H. the Z. B. Shut down computers that store data from other computers before these other computers. These data losses generally lead to a financial loss for the operator of the corresponding IT network.

So-called uninterruptible power supplies (UPS) with integrated Simple Network Management Protocol (SNMP) hardware can be used to avoid a random shutdown. A common UPS has one or more batteries with a certain battery capacity and detects a power failure to ensure the power supply of a connected computer within the existing battery capacity. In particular, such a UPS informs every computer, which is also equipped with SNMP hardware and works in the same network segment as the UPS, about a power failure. An SNMP software that is universal for the corresponding network can then be configured with corresponding effort so that when a long-term occurs In the event of a power failure, in the event that the available battery capacity is insufficient to ensure the power supply to the computers, appropriate steps are triggered, such as a shutdown of the computers. This creates a multi-stage process by which the entire network, equipped with SNMP hardware and correctly configured, is shut down in a controlled manner.

A major problem here is that only networks that are fully equipped with SNMP hardware can be shut down in a controlled manner without data loss. Often, however, there is no communication to connected networks with less high-quality equipment, i.e. Networks that do not have SNMP-capable hardware.

Furthermore, SNMP-capable hardware in the configuration described is expensive, in particular since each network equipped with SNMP-capable hardware and software requires a specialist for installation and support. A significant administrative effort arises especially in the event that a corresponding network has to be restarted after a long-lasting power failure.

Another major disadvantage in SNMP systems is that security gaps easily creep into such systems because the SNMP-capable software allows a so-called hacker to obtain any information about the network as well as all computers integrated in this network. Access to the information within the SNMP network is so easy because complex network systems are often systematized by the corresponding network administrators, for example based on geographical locations, department names or projects that are currently being worked on. However, such systems are easy to understand, especially for experienced hackers.

Another possibility is to equip each computer with a UPS that contains a notification function to shutdown the corresponding computer. The individual computers can then be configured in this way by presetting a time remaining for a corresponding computer until the shutdown. figures that computers of higher importance are shut down after computers of lower importance.

A major problem in such a configuration is that this configuration is only suitable for simple networks, since every computer must be equipped with a notifying UPS. If, for example, an existing network is enlarged and thus becomes more complex, the probability increases that errors occur when the existing configuration is revised.

Another option is to equip each computer and each network peripheral device with a UPS without a notification function that acoustically reports the power failure. The respective user of the computer or a network administrator is then required to shut down the computers on the network in the correct order.

A major disadvantage of this option is that there is only a limited amount of time to shutdown the computers due to the limited battery capacity of the UPS. Furthermore, the administrative effort is correspondingly high, which means that the shutdown can easily lead to irreparable errors or data loss.

Starting from the prior art described above, it is the object of the present invention to provide a method which enables data losses in complex EDP networks to be avoided in the event of longer-lasting power failures and which minimizes the administrative effort when restarting these EDP networks after prolonged power failures.

The above object is achieved by the subject matter of claims 1 and 18.

Preferred embodiments of the present invention are the subject of the dependent claims. In particular, the above object is achieved by a method for controlling a multiplicity of computers in the event of a power failure, the multiplicity of computers being connected to one another via at least one network, a first controlling computer being determined from the multiplicity of computers, which occurs when a Power failure performs a specified control process, an order of the plurality of computers is specified to ensure an orderly execution of the control process, and the first controlling computer detects the occurrence of a power failure and via the at least one network the control process regarding the plurality of computers in the specified Executes order in order.

In particular, the method according to the invention is used in already existing network topologies and can be used to inform physically distant networks or computers of a power failure of a corresponding network in order to request these networks or computers to initiate appropriate measures.

The control process preferably represents sending an electronic message. According to a variant, the electronic message can represent a request to shutdown or trigger a controlled shutdown or startup of a computer from the plurality of computers. According to another variant, the electronic message can represent a request to terminate communications, a controlled termination of such communications or a request to resume interrupted communications of a computer from the plurality of computers via the at least one network.

The electronic message is preferably sent via a protocol based on network protocols in the third layer in the ISO / OSI layer model. According to another variant, the electronic message can be an email.

The at least one network is preferably divided into a plurality of levels and each level is divided into a plurality of groups, with a computer being divided into a group with regard to the importance or quality of the data to be processed by the corresponding computer. Furthermore, the division into groups is preferably carried out in such a way that no dependencies within a group between computers. In this case, the control process is carried out in a level-oriented manner, the control process preferably being carried out simultaneously with respect to all computers which are divided into a common group.

Each computer of the plurality of computers preferably has a working memory in which a software program is stored when the computer is switched on, which software enables the corresponding computer to be shut down, the controlling computer sending at least one electronic message to each computer of the plurality of computers, to request the corresponding computer to execute the assigned software program. The electronic message is preferably sent using a protocol based on network protocols in the third layer in the ISO / OS I layer model. According to another variant, the electronic message is an email.

In particular, the software program can comprise a plurality of modules, each module being assigned to a specific application program and enabling this application program to be shut down. A corresponding module can be stored in the working memory of the corresponding computer when the assigned application program is started.

Furthermore, the plurality of computers can comprise at least one mobile computer, an electronic message being sent to the at least one mobile computer in order to inform the at least one mobile computer of the detected power failure. The electronic message can be sent using a protocol based on network protocols in the third layer in the ISO / OS I layer model. The at least one mobile computer can be a laptop or a notebook, the electronic message being, for example, an email. The at least one mobile computer can also be a mobile phone, the electronic message being, for example, an SMS.

Furthermore, a second controlling computer can be determined from the plurality of computers, which executes the control process in the event that the first controlling computer fails. According to a particularly preferred embodiment, the above object is achieved by a method for shutting down a plurality of computers in the event of a power failure, the plurality of computers being connected to one another via at least one network and each computer having a working memory, a first controlling computer being composed of the plurality of computers that controls the shutdown of all computers is determined, an order in which the computers are shut down to allow an orderly shutdown of the plurality of computers, a software program is stored in the memory of each computer each time it is turned on , which enables the shutdown of the corresponding computer, and the first controlling computer detects the occurrence of a power failure and via the at least one network causes the stored software programs of the plurality of computers in the fixed order to shutdown the large number of computers in an orderly fashion.

According to the present invention, short-term power failures are compensated for by a commercially available UPS. In the event of longer-lasting power failures, the large number of computers in the at least one network are arranged and shut down according to the importance of the computer in order to avoid serious data losses. The controlling computer is preferably used to store data material for the large number of computers and can only be shut down when a so-called OK message has been sent to the controlling computer from all the computers in the large number of computers. In particular, the battery capacities of the connected UPSs are designed in such a way that the power supply of each computer is ensured until subordinate computers and the corresponding computer itself have been shut down. In particular, an order is defined in which the individual computers are shut down, this order taking into account the importance of the individual computers.

The software program preferably comprises a large number of modules, each module being assigned to a specific application program and enabling this application program to be shut down. These modules can be saved in the working memory of the respective computer when the assigned application program is started. Further preferred embodiments of the present invention are explained in more detail with reference to the following figures. The figures contain:

characters

1 to 3 illustrations of different levels and groups of an exemplary network;

4 shows a flowchart to illustrate the sequence of preparatory steps when using the method according to the invention; and

Fig. 5 is a flow chart illustrating a shutdown of a plurality of computers according to the invention.

1 shows an illustration of an exemplary network level that represents a corporate headquarters. The corporate headquarters is divided into three exemplary groups: a development group 1, a communication server group 2 and a communication connection group 3. The development group 1 and the communication connection group 3 are connected to one another and to a global server via a hub 8. Center connected. An exemplary global server center is illustrated in FIG. 2 and is explained in more detail below. The communication connection group 3 is connected, for example, to the communication server group 2 via a so-called external firewall, this external firewall serving in particular to protect against so-called hackers who could, for example, gain access to the global server center via the Internet.

Development group 1 comprises a large number of computers and peripheral devices, such as, for example, a development computer 4 referred to as Devel II and a printer 5 referred to as duplex laser printer. The large number of computers and peripheral devices are preferably connected to one another by means of an Ethernet LAN, as well as via this Ethernet LAN with Hub 8.

The communication server group 2 has a large number of computers and peripheral devices, such as a mail computer referred to as an e-mail server 6. This large number of computers and peripheral devices are preferably also connected to one another via an Ethernet LAN and enable a connection to the Internet by means of suitable devices.

The communication connection group 3 has a multiplicity of computers and peripheral devices, such as, for example, a mail computer 7 referred to as an internal e-mail server and a router 9 referred to as FiliRT-1. This multiplicity of computers and peripheral devices is preferably also available Ethernet LAN in connection with one another and, by means of suitable devices, enables in particular the connection of the corporate central level to other levels, such as a partner company or a branch of the company. An exemplary branch is illustrated in FIG. 3 and is explained in more detail below.

When operating the exemplary corporate central level, a server computer suitable for this purpose, which is preferably located in the global server center, e.g. B. data and software programs for the individual groups 1, 2, 3 provided. The development computer 4 can call and execute such a software program in order to generate data which in turn are stored on the server computer. Furthermore, a user of the development computer 4 could write an e-mail, which is temporarily stored on the mail computer 6 and could be sent via the appropriate devices and the Internet. In addition, data from the server computer of the global server center could be sent to a corresponding recipient computer via router 9 for processing in the branch.

FIG. 2 shows an illustration of an exemplary network level, which represents a global server center, to which the individual groups from the corporate central level according to FIG. 1 could be connected via the hub 8. This global server center includes various server computers, e.g. B. a primary server 10, peripheral devices and at least one central database, which are preferably connected to one another via a Fiber Distributed Data Interface (FDDI) and via a suitable bridge and the hub 8 from FIG. 1 to the various groups or computers in can provide data and software programs at different network levels. 3 shows an illustration of an exemplary network level representing a branch. This branch represents a branch of the company, the company's central level is illustrated in FIG. 1.

The branch level is connected to the corporate central level through a router called FiliRT-II and includes a variety of computers and peripherals, such as an on-site customer service laptop 12, which are divided into different groups. For reasons of clarity, only a single branch group 11 is shown in FIG. 3, which is connected via a hub to corresponding communication devices and the router, the hub also being able to be used as a connection node to further groups on the branch level. As indicated in FIG. 3, the hub could, for example, represent a connection to a local server center, this local server center preferably representing a separate group.

In the following, with reference to FIGS. 1 to 3, a short example is used to explain how, in the event of a power failure, by executing a control process according to the invention in accordance with a preferred embodiment of the present invention, data loss in the exemplary network can be avoided. In the example described, the control process represents an orderly shutdown of the network computers by a controlling computer, i.e. the controlling computer sends electronic messages to the network computers representing shutdown requests or a triggered shutdown trigger. In accordance with other embodiments of the present invention, the control process may be sending messages representing startup of the network computers, requests to terminate communications, controlled termination of such communications, or requests to resume interrupted communications from the network computers over the network.

According to a particularly preferred embodiment of the present invention, all actions of the controlling computer are carried out in a level-oriented and group-based manner, so that in the following an indication of a communication of the controlling the computers with a particular computer in a group is tantamount to communicating with each computer in the corresponding group at the same level.

Short-term power failures are preferably compensated for by commercially available uninterruptible power supplies (UPS). In the event of prolonged power failures, the individual computers are shut down in order of their importance, i. H. measured by the quality of the data to be processed by the individual computers. The controlling computer is only shut down when a confirmation message has arrived from all connected computers, which confirms a successful shutdown of the corresponding computer. This confirmation message is referred to below as an OK message and is preferably sent to the controlling computer in electronic form via a separate protocol based on network protocols in the third layer in the ISO / OSI layer model. According to other embodiments of the present invention, this confirmation message can be sent as an email or SMS.

The following explanations describe the case in which a UPS is activated in the event of a power failure and thus the controlling computer, in the present example the server computer 10, detects the power failure and brings about an orderly shutdown of all computers in the network.

In a first step, the server computer 10 prompts the mail computer 6 to shut down and sets a period of 5 minutes, for example. Preferably, not only the mail computer 6 is requested to shut down, but all computers that are assigned to the same group or level as the mail computer 6. Furthermore, the server computer 10 sends a notification message, in the form of an electronic message, to the customer service computer 12 in order to notify the latter of this, for example to terminate accesses to a central database within 5 minutes. Finally, the server computer 10 sends a notification message to the development computer 4 and requests the development computer 4 to end the current work and to shut down. The corresponding computers process the messages received and activate the software programs stored in their working memory for shutdown. These software programs can be stored locally on a hard disk of the corresponding computer, as well as in the server computer 10, the corresponding computers preferably retrieving the software programs from the server computer 10 when they are switched on and storing them in their working memory.

According to a preferred embodiment of the present invention, each software program for shutting down the computer comprises a plurality of modules, each module being provided for shutting down an associated application program and being stored in the working memory of the corresponding computer when the application program is started. The individual modules can also be stored either locally on a hard disk of the corresponding computer, as well as in the server computer 10.

In the event that the software programs or modules are stored in a memory of the server computer 10, they can preferably be made available in the context of so-called "floating licenses". In particular, a license program can be provided for this case, which automatically orders new licenses of the software program in the event that insufficient "floating licenses" are available.

The mail computer 6 now preferably immediately ends all service interfaces and either waits for the current services, such as sending e-mail and receiving e-mail, to finish until the 5 minutes set for it have passed or all services have ended. When all services of the mail computer 6 have ended, the mail computer 6 is shut down and sends a corresponding OK message to the server computer 10 for confirmation. Customer service computer 12 and development computer 4 each display a message on their screens alerting the user that a power failure has occurred at the corporate central level.

In a further step, the server computer 10 sends a renewed notification message to the on-site customer service computer 12 in order to inform it thereof, that all database access to the central database is interrupted. In order to interrupt the database accesses, the server computer 10 sends the router 9 a shutdown request which is connected, for example, to a period of 30 seconds. The router 9 then terminates the connection to the corporate central level and also shuts down. The successful shutdown of the router 9 is confirmed by the server computer 10 by sending a corresponding OK message.

In a further step, the server computer 10 sends a final warning to the development computer 4 in order to request that it be shutdown. In the event that the user of the development computer 4 continues to work, this can lead to data loss when the server computer 10 is shut down, so that further work on the development computer 4 will take place at the risk of the user. Preferably, such a shutdown refusal must be authorized by a system administrator, for example. Therefore, according to another embodiment of the present invention, in the event that the user of the development computer 4 is not ready to shut down his computer, this shutdown can be initiated and performed by the server computer 10 without the user of the development computer 4 to enable this to be prevented. In the event that the user of the development computer 4 observes the last notice of the server computer 10 and shuts down his computer, a successful shutdown is again confirmed by sending a corresponding OK message to the server computer 10.

When all computers in all groups have been shut down and server computer 10 has received the appropriate OK messages, the orderly shutdown is complete and server computer 10 can be shut down.

According to a particularly preferred embodiment of the present invention, the shutdown process can be ended after a predetermined time, the server computer 10 shutting down the computers of all groups which have not yet shutdown as part of a time slice model. As an alternative to this, the server computer 10 can shut itself down after the predetermined time, with loading users of computers that are not shut down can continue to work at their own risk.

According to a further particularly preferred embodiment of the present invention, the shutdown process can also be terminated in an orderly manner as soon as the power supply to all computers is ensured again.

To clarify the communication described above and the actions performed by the controlling computer and the large number of computers in the event of a detected power failure, these are summarized in the following table:

FIG. 4 shows a flowchart to illustrate the sequence of preparatory steps when using the method according to the invention for shutting down a large number of computers in the event of a power failure.

In a first step 13, different levels are defined within a network and different computers are combined into individual groups in the respective levels. Two such network levels are shown in FIGS. 1 to 3: the corporate central level and the branch level. Several groups are illustrated at the corporate central level, for example the global server center and the development group.

In a further step 14, a controlling computer is determined from the large number of computers within the entire network, which initiates and controls a shutdown of all computers in the network when a power failure is detected by means of a UPS, and particularly preferably also physically distant networks or computers from one power failure occurs, so that these physically distant networks are informed, for example, in the event of the failure of another network or another network level, and appropriate measures can be initiated. Step 15 determines the order in which the computers on the network are shut down. This sequence preferably depends on the importance and location of individual computers of the multiplicity of computers and is preferably matched precisely to the time required by software routines used for shutting down in accordance with the capacity of individual UPSs required. Furthermore, the computers can be set in such a way that an orderly switching on or starting up of the entire network is also possible without causing an overload of the corresponding power network.

A corresponding software program for controlling the orderly shutdown of all computers is stored in the specific, controlling computer and associated software programs for shutting down the individual computers from the multiplicity of computers are each stored in their working memory when the individual computers are switched on.

FIG. 5 shows a flow diagram to illustrate an orderly shutdown of computers in a network.

In the event that a power failure 16 occurs, this power failure is detected in step 17 by the controlling computer determined according to FIG. 4, whereupon a software program for the orderly shutdown of the plurality of computers is executed in the controlling computer.

For this purpose, in step 18, all computers in the network are asked to shutdown in the order specified in FIG. 4, the computers being able to be prompted to shutdown immediately, or initially only receiving notifications, for example, which inform about the power failure. These notifications are preferably sent in the form of electronic messages via a protocol based on network protocols in the third layer in the ISO / OS I layer model. Furthermore, the corresponding computer can be requested to shutdown in a predetermined period of time by means of these notifications. According to a further variant, the controlling computer can send e-mail or SMS messages to computers, which is particularly advantageous in the case that mobile computers such as notebooks or laptops, for example, via the Internet with the controlling computer or any server computer Exchange data in the affected network. Thus, these mobile computers can be informed, for example, of performing maintenance work on the power supply of the controlling computer in the event that they cannot be reached directly by the controlling computer via the network concerned.

In step 19, a software program stored in the working memory of each computer is executed for shutting down the corresponding computer. In this case, the execution of the software program can be initiated by a user of the corresponding computer, or by the controlling computer, in the event that the corresponding computer has not been shut down in a time period set by the controlling computer. Computers whose users wish to continue working at their own risk preferably still send an appropriate OK message to the controlling computer to confirm that the information message regarding the power failure has been received, but the user is not ready to shut down. Such refusal to shutdown must preferably e.g. be authorized by an appropriate system administrator.

After all of the computers have been shut down and OK messages have been sent to the controlling computer, the controlling computer is shut down.

According to a further variant, the controlling computer is shut down after a predetermined period of time, even in the event that not all of the computers have been shut down.

An advantage of the method according to the invention for shutting down or restarting computers in a network is that no system errors can occur due to the orderly shutdown of the large number of computers, so that the administrative effort when starting up or shutting down the network can be significantly reduced, which usually brings considerable cost advantages. Another advantage is that the automation of the routines and order in an order based on the importance and location of the individual computers enables the capacity of the necessary UPSs to be precisely matched to the time required for the corresponding routines, so that in particular low-cost UPSs with low capacity can be used. This necessary capacity is also reduced by the fast response time when shutting down, which is due to the storage of the software programs in the working memory of the respective computer.

Another advantage is that standard protocols can be used when using the method according to the invention, so that the communication used can be flexibly signed or encrypted, depending on the requirements of the application environment.

Another important advantage of the system according to the invention is its high flexibility, since new computers can be integrated without any problems, as can portable or mobile computers such as laptops or notebooks.

In particular, the method according to the invention only uses existing network structures or existing infrastructures, which ensures that the method according to the invention is very easy to use.

Claims

Claims:

1. A method for shutting down a plurality of computers (4, 6, 9, 10, 12) in the event of a power failure (16), the plurality of computers (4, 6, 9, 10, 12) being connected to one another via at least one network Connection is established and each computer (4, 6, 9, 10, 12) has a working memory,

characterized in that

a first controlling computer (10) from the plurality of computers (4, 6, 9, 10, 12) is determined (14), which controls the shutdown of all computers (4, 6, 9, 10, 12),

a sequence is determined (15) in which the computers (4, 6, 9, 10, 12) are shut down in order to enable an orderly shutdown of the plurality of computers (4, 6, 9, 10, 12),

a software program is stored in the working memory of each computer (4, 6, 9, 10, 12) each time it is switched on and enables the corresponding computer (4, 6, 9, 10, 12) to be shut down, and

the first controlling computer (10) detects the occurrence of a power failure (16) and, via the at least one network, causes the stored software programs of the plurality of computers (4, 6, 9, 10, 12) to be executed in the specified order (19) to shutdown the plurality of computers (4, 6, 9, 10, 12) in an orderly manner.

2. The method according to claim 1, characterized in that the software program comprises a plurality of modules, each module being assigned to a specific application program and enabling the shutdown of this application program.

3. The method according to claim 2, characterized in that a corresponding module is stored in the working memory of the corresponding computer (4, 6, 9, 10, 12) each time the associated application program is started.

4. The method according to any one of claims 1 to 3, characterized in that at least one second controlling computer from the plurality of computers (4, 6, 9, 10, 12) is determined (14), which shuts down all computers (4, 6, 9, 10, 12) controls in the event that the first controlling computer (10) fails.

5. The method according to any one of claims 1 to 4, characterized in that the at least one network is divided into at least one level (13).

6. The method according to claim 5, characterized in that each level is divided into at least one group (1, 2, 3, 11) (13).

7. The method according to claim 6, characterized in that a division of a computer (4, 6, 9, 10, 12) into a group (1, 2, 3, 11) with respect to the quality of the computer (4 , 6, 9, 10, 12) data to be processed.

8. The method according to any one of claims 5 to 7, characterized in that the shutdown of the plurality of computers (4, 6, 9, 10, 12) is level-oriented.

9. The method according to any one of claims 5 to 8, characterized in that the shutdown of all computers (4, 6, 9, 10, 12), which are divided into a common group (1, 2, 3, 11), takes place simultaneously ,

10. The method according to any one of claims 1 to 9, characterized in that the controlling computer (10) sends at least one electronic message to each computer of the plurality of computers (4, 6, 9, 10, 12) to the corresponding computer ( 4, 6, 9, 10, 12) to execute the assigned software program.

11. The method according to claim 10, characterized in that the transmission of the electronic message takes place via a protocol based on network protocols in the third layer in the ISO / OSI layer model.

12. The method according to claim 10, characterized in that the electronic message is an email.

13. The method according to any one of claims 1 to 12, characterized in that the controlling computer (10) causes the orderly shutdown of the plurality of computers (4, 6, 9, 10, 12) by activating the corresponding software programs.

14. The method according to any one of claims 1 to 9, characterized in that the plurality of computers (4, 6, 9, 10, 12) comprises at least one mobile computer (12), the non-mobile computers (4, 6, 9, 10) are shut down in an orderly manner and an electronic message is sent to the at least one mobile computer (12) in order to inform the at least one mobile computer (12) of the detected power failure (16).

15. The method according to claim 14, characterized in that the transmission of the electronic message takes place via a protocol based on network protocols in the third layer in the ISO / OSI layer model.

16. The method according to claim 14, characterized in that the at least one mobile computer (12) is a laptop or a notebook and that the electronic message is an e-mail.

17. The method according to claim 14, characterized in that the at least one mobile computer (12) is a mobile phone and that the electronic message is an SMS.

18. A method for controlling a plurality of computers (4, 6, 9, 10, 12) in the event of a power failure (16), the plurality of computers (4, 6, 9, 10, 12) being connected to one another via at least one network Connection is established

characterized in that

a first controlling computer (10) from the plurality of computers (4, 6, 9, 10, 12) determining (14) that executes a defined control process when a power failure (16) occurs,

an order of the plurality of computers (4, 6, 9, 10, 12) is determined (15) in order to ensure an orderly execution of the control process, and

the first controlling computer (10) detects the occurrence of a power failure (16) and uses the at least one network to carry out the control process with regard to the plurality of computers (4, 6, 9, 10, 12) in an orderly fashion.

19. The method according to claim 18, characterized in that the control process represents sending an electronic message.

20. The method according to claim 19, characterized in that the electronic message represents a request to shutdown or trigger a controlled shutdown or startup of a computer from the plurality of computers (4, 6, 9, 10, 12).

21. The method according to claim 19, characterized in that the electronic message is a request to terminate communications, a controlled termination of such communications or a request to resume interrupted communications of a computer from the plurality of computers (4, 6, 9, 10, 12 ) represented by at least one network.

22. The method according to any one of claims 19 to 21, characterized in that the transmission of the electronic message via a protocol based on network protocols takes place in the third layer in the ISO / OSI layer model.

23. The method according to claim 19, characterized in that the electronic message is an email.

24. The method according to any one of claims 18 to 23, characterized in that the at least one network is divided into at least one level (13).

25. The method according to claim 24, characterized in that each level is divided into at least one group (1, 2, 3, 11) (13).

26. The method according to claim 25, characterized in that a division of a computer (4, 6, 9, 10, 12) into a group (1, 2, 3, 11) with respect to the quality of the computer (4 , 6, 9, 10, 12) data to be processed.

27. The method according to any one of claims 24 to 26, characterized in that the control process is carried out level-oriented.

28. The method according to any one of claims 24 to 27, characterized in that the control process with respect to all computers (4, 6, 9, 10, 12) which are divided into a common group (1, 2, 3, 11), simultaneously is performed.

29. The method according to claim 18, characterized in that

each computer of the plurality of computers (4, 6, 9, 10, 12) has a working memory in which a software program is stored when the computer is switched on, which software shuts down the corresponding computer (4, 6, 9, 10, 12 ) enables, and

the controlling computer (10) sends at least one electronic message to each computer of the plurality of computers (4, 6, 9, 10, 12) to request the corresponding computer (4, 6, 9, 10, 12), the associated software Program.

30. The method according to claim 29, characterized in that the transmission of the electronic message via a protocol based on network protocols takes place in the third layer in the ISO / OSI layer model.

31. The method according to claim 29, characterized in that the electronic message is an email.

32. The method according to any one of claims 29 to 31, characterized in that the software program comprises a plurality of modules, each module being assigned to a specific application program and enabling the shutdown of this application program.

33. The method according to claim 32, characterized in that a corresponding module is stored in the working memory of the corresponding computer (4, 6, 9, 10, 12) each time the associated application program is started.

34. The method according to claim 18, characterized in that the plurality of computers (4, 6, 9, 10, 12) comprises at least one mobile computer (12) and an electronic message is sent to the at least one mobile computer (12), to inform the at least one mobile computer (12) of the detected power failure (16).

35. The method according to claim 34, characterized in that the sending of the electronic message takes place via a protocol based on network protocols in the third layer in the ISO / OSI layer model.

36. The method according to claim 35, characterized in that the at least one mobile computer (12) is a laptop or a notebook and that the electronic message is an e-mail.

37. The method according to claim 36, characterized in that the at least one mobile computer (12) is a mobile phone and that the electronic message is an SMS.

38. The method according to any one of claims 18 to 37, characterized in that a second controlling computer from the plurality of computers (4, 6, 9, 10, 12) is determined (14) which carries out the control process in the event that the first controlling computer (10) fails.