Method for booting a computer system, and computer system
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- EATON INTELLIGENT POWER LTD
- Filing Date
- 2026-01-28
- Publication Date
- 2026-08-06
Smart Images
Figure EP2026052117_06082026_PF_FP_ABST
Abstract
Description
[0001] P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0002] 1
[0003] Description
[0004] METHOD FOR BOOTING A COMPUTER SYSTEM, AND COMPUTER SYSTEM
[0005] The present disclosure relates to a method for booting a computer system. The present disclosure further relates to a computer system.
[0006] One obj ect to be achieved is, inter alia, to specify an improved method for booting or restarting a computer system which in particular allows for an automatic boot in particular of a complex computer system. A further obj ect to be achieved is to specify an improved computer system on which the method can be carried out .
[0007] These obj ects are achieved, inter alia, by a method comprising the features of independent claim 1 and by a computer system comprising the features of independent claim 16, respectively. Advantageous embodiments and further developments are the subj ect of the dependent claims .
[0008] In at least one embodiment of the method for booting or restarting a computer system, the computer system comprises at least one server, a power supply system configured to supply power to the server, and an infrastructure control unit integrated in the power supply system.
[0009] The server is preferably an electronic device and may comprise a computational component, a storage component and / or a network component . In particular, the server may carry out computational functions, storage functions and / or network functions of the computer system during operation. For example, the server is configured to carry out more thanP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0010] one of these functions . This means in particular that the server may be highly converged regarding its functions in the computer system. "Highly converged" in particular means that a plurality of functions are realized in one component, for example in one server, wherein the functions are interconnected and comprise a plurality of dependencies . Thus, the functions may not operate independently from another but are connected and linked to one another in a complex manner .
[0011] Consequently, booting such a complex server or computer system requires a specific sequence of instructions and commands to configure and synchronize the functions in accordance with the plurality of dependencies .
[0012] For example, these functions or functional components of the server such as the computational function or component, the storage function or component and / or the network function or component can be virtually emulated on the server, for example by software . It is also possible that at least one of the computational component, the storage component and / or the network component is realized on hardware level .
[0013] For example, the computer system may comprise at least two servers functionally connected to each other . This means in particular that a plurality of functional connections is provided between the servers, connecting different functional components of the complex infrastructure . For example, each of the computational component, storage component and / or network component of each server is functionally connected to each of the computational component, storage component and / or network component of another server . In this way the plurality of servers forms a complex infrastructure according to the present example . In particular, the computer system may be a complex system.P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0014] 3
[0015] The computer system may be part of or may be a data center, for example . For example, the plurality of servers may include SSH servers and servers providing web services .
[0016] The functional connection between the servers may be direct and / or indirect . In case of a direct connection, the servers may exchange data directly via a communication channel, which may be formed by a wireless connection and / or by a cable . In case of an indirect connection, the servers may communicate with each other via a third element, for example the power supply system and / or the infrastructure control unit . It is also possible that the third element is a network switch or a router .
[0017] The power supply system is preferably configured to supply power to the server or to each of the servers . For example, the power supply system comprises at least one power outlet, wherein each outlet is connected to one server . For example, the power supply system is externally supplied with power by a power source such as the public power grid. In particular, the power supply system is a central power supply for the whole computer system. This means that all components and units of the computer system are directly or indirectly supplied with power by the power supply system.
[0018] In particular, the power supply system may comprise a plurality of power devices . For example, the power supply system may comprise an uninterruptable power supply and / or at least one power distribution unit . The uninterruptable power supply may operate as a central power supply to the computer system connected to an external power source . The uninterruptable power supply may be configured to supply power to the at least one power distribution unit .P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0019] Preferably, each server is connected to exactly one power distribution unit for power supply.
[0020] The infrastructure control unit may be integrated in any of the power devices of the power supply system. For example, the infrastructure control unit is integrated in the uninterruptable power supply. Alternatively, the infrastructure control unit may be integrated in one of the at least one power distribution unit .
[0021] The infrastructure control unit may be part of a computational unit or a communication unit of the power supply system or a power device of the power supply system. The computational unit and / or the communication unit may be a microprocessor implemented on, for example, a network card. This means that the infrastructure control unit may be an individual electrical component or may be integrated in an electrical component .
[0022] It is also possible that the infrastructure control unit comprises an engine . In this case, the infrastructure control unit may be implemented as a software such as an embedded software that is executed on the power supply system or a power device of the power supply system. Preferably, the infrastructure control unit is not an individual component that is only functionally connected or part of to the power supply system.
[0023] In the present embodiment, the method for booting or restarting the computer system comprises a step of detecting a trigger event . The trigger event in particular initiates booting or restarting the computer system.P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0024] 5
[0025] For example, if the computer system is previously shut down due to a power shortage, the trigger event may be the reestablishment of the external power supply to the power supply system. If the computer system has been shut down prior to the method described herein, the method described herein in particular is a method for restarting the computer system.
[0026] In another example, the trigger event is exceeding or falling below a threshold of an environmental parameter such as a temperature . In this case, at least a part of the computer system, for example the server, has been shut down due to exceedance of a temperature threshold, for example . It is possible that in this case the power supply system is not shut down but is monitoring the environmental parameter, i . e . the ambient temperature . Once the temperature falls below the threshold, the shut down parts of the computer system are botted or restart .
[0027] In a further method step, the server is booted by executing a boot protocol, wherein at least some instructions of the boot protocol are provided by the infrastructure control unit . Furthermore, execution of the boot protocol is monitored by the infrastructure control unit . In particular, the server and / or other components of the computer system are connected to the infrastructure control unit so that the infrastructure control unit can control, monitor and orchestrate booting of the server and / or the other components of the computer system.
[0028] Preferably, the instructions of the boot protocol provided by the infrastructure control unit comprise at least mandatory start instructions for booting the server . This means inP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0029] particular, that at least such instructions that are essential for booting the server are provided by the infrastructure control unit and in particular not by another component of the computer system. For example, a plurality of instructions of commands are necessary to set up, configure and / or synchronize the server or functions provided by the server . At least these mandatory start instructions, which are included in the boot protocol, are provided by the infrastructure control unit .
[0030] For example, the boot protocol comprises N instructions or commands, and M instructions or commands are provided by the infrastructure control unit, wherein N and M are a natural numbers greater than 1 and M is smaller or equal to N. If M is equal to N, all instructions or commands of the boot protocol are provided by the infrastructure control unit .
[0031] If M is smaller than N, K instructions or commands may be provided by a component or unit of the computer system different from infrastructure control unit, for example by a server, wherein K is a natural number and K = N - M. In this case, at least some of the M commands and instructions are provided and executed prior to the K instructions and commands and instructions . In particular, these M instructions and commands comprise the mandatory start instructions for booting the at least one server .
[0032] In an optional method step, the power supply system and the infrastructure control unit are booted prior to the server . In particular, the power supply system and the infrastructure control unit are booted in response to the trigger event . In this case, the trigger event is, for example, a restorage of an external power supply. For example, the computer systemP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0033] 7
[0034] and in particular the power supply system was shut down due to a power shortage and if power is restored, the power supply system and the infrastructure control unit restart .
[0035] In other cases, for example if the trigger event is exceeding or falling below a threshold of an environmental parameter, the power supply system and thus the infrastructure control unit are not completely shut down previously but monitor the environmental parameter .
[0036] In particular, the power supply system and preferably the infrastructure control unit are booted prior to the servers . Further preferably, the power supply system and the infrastructure control unit are booted prior to any other component, element or entity of the computer system. In other words, the power supply system is started first .
[0037] For example, the power supply system and the infrastructure control unit are booted sequentially. Preferably the infrastructure control unit is booted immediately after the power supply system. The infrastructure control unit is preferably booted directly subsequently to the power device of the power supply system, in which the infrastructure control unit is integrated. This means in particular, no other component of the computer system, in particular no server, is booted between the infrastructure control unit and the corresponding power device of the power supply system.
[0038] In particular, the boot protocol includes steps of starting, configuring and / or synchronizing the server . The steps of the boot protocol are in particular determined by interdependencies of the server or between the servers .P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0039] 8
[0040] For example, the boot protocol includes steps of controlling a power controller of the uninterruptible power supply included in the power supply system to turn on or off at least one load segment to deliver electricity to other components of the computer system. In particular, by the load segment the uninterruptible power supply provides power and / or electricity to other components of the computer system, which may include the server or other power devices of the power supply system such as the power distribution unit, for example .
[0041] It is also possible that the boot protocol includes steps of connecting to a computational unit of the power distribution unit included in the power supply system to control a power controller of the power distribution unit to turn on or off at least one power outlet of the power distribution units to provide electricity to other components of the computer system. In particular, by the load segment the uninterruptible power supply provides power and / or electricity to other components of the computer system, which may preferably include the server or other parts of the computer system.
[0042] In a specific example, the uninterruptible power supply provides power and / or electricity to the power distribution unit via the load segments during operation. Further, in the specific example, the power distribution unit provides power and / or electricity to the at least one server via the power outlets . The infrastructure control unit in particular is configured to control the power and / or electricity provided by the load segments and the power outlets by connecting to the power controller of the uninterruptible power supply andP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0043] 9
[0044] the computational of the power distribution unit, respectively .
[0045] According to one example, due to the functional connection between the servers it is necessary that a first server is configured before a second and a third server can be fully configured since these servers require data that are provided by the first server . Subsequently, the second sever may be configured upon receiving the required data from the fully configured first server . Subsequently to configuration of the second server, the first and second servers may be required to synchronize to provide the required data to the third server for proper configuration. This means that only subsequent to synchronization of the first and second servers, the third server may be configured. Thus, in this illustrative example, the boot protocol comprises a plurality of steps or instructions by which the first, second and third servers may be properly configured and booted. In particular, some of these instructions are only provided after other steps initiated by the boot protocol are terminated, such as configuration of the first server or synchronization of the first and second servers .
[0046] According to another example, the computer system may comprise essentially one server, which is configured to carry out a plurality of functions . These functions may be implemented on the server in a complex and inter-connected manner . That is, the functions may have a plurality of dependencies from each other so that sequential and interdependent actions are required to correctly boot the server . In this case, the boot protocol preferably includes explicit instructions to carry out this sequence of actionsP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0047] 10
[0048] appropriately so that a correct boot process can be carried out .
[0049] It is also possible that the computer system may comprise a plurality of servers that are highly dependent from each other and at the same time, each of the servers has a complex infrastructure . In this case, the boot protocol comprises a plurality of instructions that have to be carried out in a specific sequence to correctly boot the computer system.
[0050] The method described herein is based on the following technical considerations . Compared to common simple computer systems, where there are little or no dependencies between individual components, complex computer systems, such as data centers, offer a number of tiers involved in the computer system. Thus, operational and maintenance costs can be reduced, for example .
[0051] In certain events such as a power shortage or overheating or the like, the computer system has to be shut down at least partially to protect the computer system. If the potentially harmful event is overcome, i . e . power restored, for example, the computer system has to be restarted. However, due to the complex infrastructure of the computer system, a plurality of dependencies between the individual components and severs of the computer system are present as illustrated above .
[0052] Therefore, restarting or booting the computer system is a challenging task, which commonly requires external interference, for example by a human technician and / or external relays that coordinate and monitor the restarting procedure . Hence, shutting down and restarting such a complex computer system entail inconvenience and significant costsP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0053] including downtime of the computer system, human working hours, travel costs and so on.
[0054] The method described herein is based on the idea of integrating the infrastructure control unit configured to control and coordinate the booting or restart of the computer system via the boot protocol in the power supply system, thus in the computer system. This allows, inter alia, to carry out an automatic boot or restart of the computer system upon detection of a trigger event, which may indicate that a power shortage is overcome .
[0055] Thus, an external interference, for example by a human technician, can be advantageously reduced or avoided.
[0056] Furthermore, additional relays that control, coordinate and monitor the restarting process can be omitted.
[0057] As a result, costs associated with a shutdown event of the computer system can be significantly reduced. Furthermore, uptime of the computer system can be increased without additional external relays, simplifying the restart procedure .
[0058] Furthermore, since the power supply system and preferably the infrastructure control unit may be booted prior to the servers and are further preferably booted directly sequentially to each other, the infrastructure control unit can efficiently control, coordinate and monitor the booting or restarting procedure of the server or the servers .
[0059] As the boot protocol as described above comprises a plurality of instructions that allow to properly start or boot the server or the servers, wherein the internal dependencies ofP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0060] - 12 -
[0061] the server and / or dependencies of the servers among each other are considered, also large-scale computer systems, which may include hundreds or thousands of servers, can be automatically booted or restarted. Moreover, since the infrastructure control unit monitors the booting procedure of the computer system, the boot protocol may be adapted to unexpected events such as malfunction and / or internal failure of certain servers .
[0062] According to a preferred embodiment, the infrastructure control unit is configured to execute a plurality of boot protocols . Depending on the trigger event, a corresponding boot protocol is executed by the infrastructure control unit .
[0063] For example, different events including, but not limited to, power shortage, overheating, smoke detection or vibration detection, may trigger a shutdown of the computer system. If these events are overcome, different boot protocols may be executed, depending on the overcome event, in order to boot or restart the computer system.
[0064] Additionally or alternatively, the trigger event may be exceeding or falling below a threshold of an environmental parameter or, for example, any change of state of the environmental parameter . For example, the environmental parameter may be temperature, a smoke concentration, humidity, vibrations . If a threshold of one of these parameters is exceeded, there may be a risk of damaging the computer system. Thus, the computer system may at least partially shut down and, if the threshold is undercut, the computer system may reboot in a secure environment . For example, the server is shut down, but the power supply unitP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0065] and thus the infrastructure control unit may be keep running for monitoring the environmental parameter .
[0066] Additionally or alternatively, the trigger event may be reaching a scheduled event for booting or restarting the computer system. For example, the computer system may be fully or partially shut down at certain times for maintenance or power saving, for example during weekends or business closing hours . In this case it is desired that the computer system is available after a certain period of time or at a certain point in time, which is preferably the scheduled event . For example, the server is shut down, but the power supply unit and thus the infrastructure control unit may be keep running for monitoring reaching the scheduled event .
[0067] It is also possible that the trigger event is presence of absence of power externally provided to the power supply system. For example, the computer system is shut down after a power shortage and is to be restarted after the power is restored. In particular, also the power supply system and the infrastructure control unit may be shut down and have to be booted before the servers .
[0068] It is further possible that the trigger event may be exceeding or falling below a threshold of a quality parameter of the power externally provided to the power supply system. For example, the quality parameters may include a frequency of the external power signal or any other property or signal parameter of the external power .
[0069] It is further possible that the trigger event may be presence or absence of an internal failure of a component of the computer system. The internal failure may be a malfunction orP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0070] - 14 -
[0071] damage of a component such as a server or a power device of the power supply system.
[0072] For example, if the computer system is to be restarted due to overcome power shortage, a first boot protocol is executed by the infrastructure control unit . According to another example, if the computer system is to be restarted due to an overheating event, a second boot protocol is executed by the infrastructure control unit .
[0073] Different overcome interrupting events may require different processes to ensure safe booting of the computer system. For example, in cases of overheating, a sequence for booting the servers may be different to ensure a stable and homogeneous temperature in the computer system to avoid the risk of reoccurrence of overheating.
[0074] Advantageously, carrying out different boot protocols according to different trigger events allows for proper booting of the computer system in a plurality of different situations .
[0075] According to a further preferred embodiment, the computer system comprises or is connected to at least one sensor, wherein the sensor detects the trigger event .
[0076] The sensor may comprise a power sensor, a temperature sensor, a humidity sensor, a smoke detector, a contact sensor, a motion sensor, a vibration detector, a clock, any external application programming interface, API, external monitored system or the like .
[0077] The sensor may be an internal or external sensor . In case of an internal sensor, the sensor is preferably integrated inP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0078] the computer system, for example integrated in a power device of the power supply system or the server . In case of an external server, the server may be connected to the computer system, for example via a data communication line .
[0079] Additionally, the sensor may be a hardware component or a software component or a mixture thereof .
[0080] For example, if the trigger event is exceeding or falling below a threshold of an environmental parameter, the sensor may be an external server that detects the environmental parameter remote to the computer system. However, it is also possible that the sensor may be an internal sensor .
[0081] If, in another example, the trigger event comprises reaching a scheduled event for booting or restarting the computer system, the system is preferably an internal server, for example a clock, that is further preferably implemented as a software .
[0082] If, according to a further example, the trigger event is presence or absence of power externally provided to the power supply system, the sensor is preferably an internal power sensor integrated in the power system, for example integrated in the uninterruptable power supply.
[0083] If, in a further example, the trigger event is exceeding or falling below a threshold of a quality parameter of the power externally provided to the power supply system, the sensor is preferably an internal sensor analyzing a power signal of the external power . The sensor may be integrated in the power supply system, in particular in the uninterruptable power supply and preferably is implemented as a software .P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0084] - 16 -
[0085] According to another example, if the trigger event comprises presence or absence of an internal failure of a component of the computer system, the sensor is an internal sensor . This internal server may be included in the power supply system, the server or another component of the computer system monitoring proper functionality of the corresponding component .
[0086] In particular, a sensor signal is generated by the at least one sensor in response to the trigger event . The trigger event may be any of the above-mentioned events but is not limited to these examples .
[0087] According to a preferred embodiment, the power supply system comprises a decision unit . The decision unit may be integrated in the power supply system, preferably in a power device of the power supply system also comprising the infrastructure control unit . Further preferably, the decision unit is integrated in the infrastructure control unit . The decision unit may be implemented as a software carried out or executed by the infrastructure control unit .
[0088] The decision unit monitors the at least one sensor signal provided by the at least one sensor and determines the presence of a trigger event depending on the monitored sensor signal .
[0089] For example, the at least one sensor provides the sensor signal as a continuous signal . In this case, the sensor may be an environmental sensor continuously providing the environmental parameter such as temperature or humidity as the sensor signal and the decision unit decides whether aP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0090] predefined threshold of such an environmental parameter is exceeded or not .
[0091] It is also possible that the sensor signal is generated in response to a specific event, for example if the trigger event is reaching of a scheduled event, the reestablishment of a power supply or a detection of an internal failure . In this case, the decision unit determines the presence of the trigger event if any sensor signal is provided.
[0092] In another preferred embodiment of the method, the infrastructure control unit comprises a feedback unit, which provides a feedback signal . Depending on the feedback signal, the infrastructure control unit adapts or terminates the boot protocol . For example, the feedback unit may comprise at least one feedback sensor, which may detect a temperature or measure an electrical power . Preferably, the feedback sensor may detect an internal failure of the computer system or monitor a grid stability of the complex infrastructure of the computer system.
[0093] By the feedback unit unexpected situations during booting of the servers may be observed, for example local overheating or grid instabilities in the server infrastructure . By generating the feedback signal and providing the feedback signal to the infrastructure control unit, the infrastructure control unit may adapt the boot protocol such that, for example, certain servers are booted later or booting of such servers is omitted, in order to reduce a risk of harming the computer system. For example, adaption of the boot protocol may be carried out based on previous monitoring data or defined criteria . It is also possible that the infrastructureP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0094] control unit terminates the booting procedure to avoid damaging of the computer system if necessary.
[0095] Furthermore, the infrastructure control unit may control the power supplied to the power supply system depending on the feedback signal . For example, the infrastructure control unit may be connected to power controllers, by which power that is coupled from the external grid to the power supply system of the computer system can be controlled. If, for example, the feedback unit detects, for example on the basis of previous monitoring data or manually defined criticality, exceedance of a power threshold, a corresponding feedback signal may be generated. In response to this feedback signal the infrastructure control unit may reduce the power or even cut the power off the computer system by means of controlling the power controllers in order to avoid damage to the computer system.
[0096] According to an embodiment of the method, the boot protocol is executed by a computational unit or communication unit of the infrastructure control unit or of the power supply system. For example, the boot protocol may be executed by a computational or communication unit of a power device of the power supply system such as the uninterruptable power supply or the at least one power distribution unit .
[0097] The computational unit and / or the communication unit may be a microprocessor implemented on, for example, a network card. This means that the infrastructure control unit may be an individual electrical component or may be integrated in an electrical component . It is also possible that the infrastructure control unit comprises an engine . In this case, the infrastructure control unit may be implemented as aP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0098] software such as an embedded software that is executed on the power supply system. Preferably, the infrastructure control unit is not an individual component that is only functionally connected to the power supply system.
[0099] According to a further embodiment of the method, the computer system comprises a plurality of servers and some servers are prioritized in the boot protocol . Depending on the application of the computer system, some of the servers may be critical for ensuring a basic state of operation of the computer system, and other servers may be of minor importance to ensuring this basic state . The other servers may, for example, provide redundant services to enhance reliability of the computer system or provide functions that are not mandatory for basic operation. Therefore, the booting of the critical servers may be prioritized over booting the other servers in the boot protocol . This advantageously allows to reduce downtime of the computer system.
[0100] According to at least one embodiment of the method, the infrastructure control unit is connected to a server control unit integrated in the at least one server and the infrastructure control unit instructs the server to execute applications, to execute configuration of server systems and applications, and / or to start nested virtual computer systems .
[0101] According to at least one embodiment of the method, the feedback unit receives signal, states and metrics of the at least one server from the server control unit, and / or signal, state and metrics from power devices of the power supply system. For example, by these signals, states and metrics, the feedback signal can be generated. As a result, the bootP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0102] - 20 -
[0103] protocol may be adapted if these signals, states and metrics show a malfunction.
[0104] It is also possible that the feedback unit and / or one of the sensors may receive signals, states and metrics from external systems . Such external system may include any system outside the computer system that may interfere with operating the computer system. For example, the signals, states and metrics from external systems may be provided by an external provider giving information about quality parameters of the external power or the like .
[0105] Furthermore, a computer system is specified. The computer system may be booted or restarted by carrying out the method described herein in accordance with one or more embodiments during operation. This means that all features disclosed for the method are also disclosed for the computer system and vice versa .
[0106] In at least one embodiment, the computer system comprises at least one server, a power supply system configured to supply power to each of the servers, and an infrastructure control unit integrated in the power supply system. The infrastructure control unit is configured to boot the servers according to a predefined boot protocol upon detection of a trigger event by providing at least some instructions of the boot protocol . The infrastructure control unit is further configured to monitor execution of the boot protocol .
[0107] The server is preferably an electronic device and may comprise a computational component, a storage component and / or a network component . In particular, the server may carry out computational functions, storage functions and / orP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0108] network functions of the computer system during operation. For example, the server is configured to carry out more than one of these functions . In particular, these functions are connected to each other defining a plurality of dependencies between these functions and / or components .
[0109] For example, the computer system comprises at least two servers functionally connected to each other . This means in particular that a plurality of functional connections are provided between the servers, connecting different functional components of the highly connected servers . For example, each of the computational component, storage component and / or network component of each server is functionally connected to each of the computational component, storage component and / or network component of another server . In this way the plurality of servers forms a complex and highly connected infrastructure . In particular, the computer system may be a complex and highly connected computer system. The computer system may be part of or may be a data center, for example .
[0110] The server is configured to be provided with power exclusively via the power supply system. For example, the power supply system comprises a plurality of power devices, for example the uninterruptable power supply and the at least one power distribution unit .
[0111] The infrastructure control unit is integrated in the power supply system or a power device of the power supply system. The infrastructure control unit may be part of a computational unit or a communication unit of the power supply system or a power device of the power supply system.P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0112] The computational unit and / or the communication unit may be a microprocessor implemented on, for example, a network card. This means that the infrastructure control unit may be an individual electrical component or may be integrated in an electrical component .
[0113] It is also possible that the infrastructure control unit comprises an engine . In this case, the infrastructure control unit may be implemented as a software such as an embedded software that is executed on the power supply system.
[0114] Preferably, the infrastructure control unit is not an individual component that is only functionally connected to the power supply system.
[0115] According to a preferred embodiment of the computer system, the computer system comprises or is connected to at least one sensor, wherein the sensor is configured to provide a sensor signal, for example in response to the trigger event . In other words, the sensor is configured to detect the trigger event . The sensor in particular may comprise at least one of a clock, an environmental sensor, a power sensor but is not limited thereto . The sensor may be an internal or external sensor .
[0116] Preferably, the infrastructure control unit comprises a decision unit configured to determine the presence of a trigger event depending on the sensor signal .
[0117] According to at least one embodiment, the power supply system comprises the uninterruptible power supply and at least one power distribution unit electrically connected to the uninterruptible power supply. Each of the uninterruptible power supply and the power distribution unit comprises aP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0118] - 23 -
[0119] computational unit connected to each other by a data connection, and one of the computational units comprises the infrastructure control unit . The data connection may be a wireless or a wire-based connection such as an ethernet connection. By the connection between the uninterruptible power supply and the power distribution unit, the infrastructure control unit can send and receive instructions or status information of the individual components . Thus, monitoring of the booting procedure of the computer system is possible .
[0120] In particular, each power distribution unit comprises at least one power outlet, wherein each power outlet may be connected to one server . For example, if the computer system comprises a plurality of servers, the servers may be arranged in server racks . For example, for each server rack, one power distribution unit is provided. This means in particular that the uninterruptible power supply provides power to the complete computer system by powering the power distribution units of each server rack, via which the servers are supplied with power . Since the uninterruptible power supply and the power distribution units are connected to one another, the power distribution in the computer system can be controlled.
[0121] Preferably, the server or each server comprises a control unit, and the infrastructure control unit is connected to the server or each server by a data connection. By this data connection the infrastructure control unit can send and receive instructions or status information of the individual servers . For example, the infrastructure control unit may provide instructions to the servers for configuration or synchronization or the like, and the servers may provideP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0122] 24
[0123] information if the corresponding process is successfully completed .
[0124] In a similar way, the servers may provide information about malfunctions of the servers . From such information, a feedback signal may be generated by the feedback unit . Thus, monitoring of the booting procedure of the servers is possible .
[0125] Further advantages and advantageous embodiments and further developments of the method and the computer system described herein will become apparent from the following exemplary embodiments shown in connection with schematic drawings . Identical elements, elements of the same kind or elements having the same effect are provided with the same reference signs in the figures . The figures and the proportions of the elements shown in the figures are not to be regarded as true to scale . Rather, individual elements may be shown exaggeratedly large for better representability and / or for better comprehensibility.
[0126] In the figures :
[0127] Figure 1 shows a schematic representation of a computer system described here according to a first exemplary embodiment;
[0128] Figure 2 illustrates a method described herein according to a first exemplary embodiment;
[0129] Figure 3 shows a schematic representation of a computer system described here according to a second exemplary embodiment .P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0130] - 25 -
[0131] Figure 1 illustrates a computer system 100 according to a first exemplary embodiment . The computer system 100 comprises a power supply system 110 and a plurality of servers 121, 122. The servers 121, 122 are supplied with power via the power supply system 110.
[0132] The power supply system 110 comprises an uninterruptible power supply 180 and a power distribution unit 112. The uninterruptible power supply 180, UPS for short, comprises a master load segment 191 and a first load segment 192. Via the first load segment 192 power can be delivered to the power distribution unit 112, PDU for short, via a power connection 10. During operation of the computer system 100, the uninterruptible power supply 180 is connected to an external power grid for power supply. Via the uninterruptible power supply 180 power from the external power supply is distributed in the computer system 100.
[0133] The power distribution unit 112 comprises a plurality of power outlets 114, 115, wherein each power outlet 114, 115 is connected to one server 121, 122.
[0134] In alternative embodiments, the computer system 100 may comprise one or more than two servers . For example, a group of servers may be collected in one server rack. The servers of each server rack are supplied with current via an individual power distribution unit 112. This means that in this case the computer system 100 comprises a plurality of power distribution units 112. Furthermore, the uninterruptible power supply 180 may comprises a plurality of load segments, each load segment may be corresponding to one power distribution unit 112.P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0135] - 26 -
[0136] The uninterruptible power supply 180 comprises a computational unit 111, which is, for example, a communication card or a network card. The computational unit 111 comprises an infrastructure control unit 151. The infrastructure control unit 151 may be an individual component such as a microcontroller integrated in the computational unit 111. It is also possible that the infrastructure control unit 151 is implemented as a software executed on the computational unit 111. The infrastructure control unit 151 is integrated in one power device of the power supply system 110 such as the uninterruptible power supply 180 or the power distribution unit 112. In the present exemplary embodiment, the infrastructure control unit 151 is integrated in the uninterruptable power supply 180.
[0137] The power distribution unit 112 comprises a further computational unit 113, which may also be a communication card or network card. The computational unit 111 of the uninterruptible power supply 180 is connected to the computational unit 113 of the power distribution unit 112 via a first data connection 5. The first data connection 5 is, for example, an ethernet connection.
[0138] Each of the servers 121, 122 is an electronic device and may comprise a computational component, a storage component and / or a network component (not shown) . This means that each of the servers may carry out computational functions, storage functions and / or network functions of the computer system 100 during operation. For example, each of the servers 121, 122 is configured to carry out more than one of these functions . This means in particular that the servers 121, 122 are interdependent regarding their functions in the computer system 100 .P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0139] - 27 -
[0140] The servers 121, 122 are functionally connected to each other . This means that a plurality of functional connections is provided between the servers 121, 122, connecting different functional components of the inter-dependent servers 121, 122. For example, each of the computational component, storage component and / or network component of each server 121, 122 is functionally connected to each of the computational component, storage component and / or network component of the other server 121, 122. In this way the servers 121, 122 form a complex, inter-dependent and highly connected infrastructure . In particular, the computer system 100 may be a complex, inter-dependent and highly connected computer system 100. The computer system 100 is a data center, for example .
[0141] As an example, in the present exemplary embodiment, the first server 121 is an SSH server and the second server 122 is a server providing web services .
[0142] Each of the servers 121, 122 comprises a control unit 141, 142. The control unit 141, 142 is in particular connected to the infrastructure control unit 151 via second and third data connections 6, 7, respectively. The second and third data connections 6, 7 may be ethernet connections .
[0143] The infrastructure control unit 151 is configured to boot the computer system 100 upon receiving a sensor signal 2 by executing a predefined boot protocol 161, as will be explained in the following in connection with Figure 2.
[0144] In the present embodiment, the boot signal 2 is an external boot signal, for example generated by an external sensor in response to a trigger event . For example, the trigger eventP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0145] is falling down below a threshold of an environmental parameter after a previous exceeding of the parameter . For example, the parameter is an ambient temperature . In this example, the computer system 100 was partially shut down due to a potential risk of overheating due to high ambient temperature . After decreasing of the ambient temperature, the computer system 100 may be completely booted or restarted.
[0146] In other exemplary embodiments, the trigger event may be a restorage of external power supply. For example, the computer system 100 was completely shut down due to a power shortage and is to be restarted after the power is restored. In this case, a sensor signal 1 may be generated as a response to the restorage of power triggering restarting of the computer system 100.
[0147] The infrastructure control unit 151 comprises a decision unit 153. The decision unit 153 monitors sensor signal 2 and determines the presence of the trigger event depending in the monitored sensor signal 2.
[0148] In the present exemplary embodiment, the sensor signal 2 as a continuous signal, indicating the ambient temperature . The decision unit 153 is configured to decide whether a predefined threshold of the ambient temperature is exceeded or not .
[0149] The infrastructure control unit 151 further comprises a feedback unit 152. The feedback unit 152 may include at least one sensor configured to monitor booting of the computer system 100. For example, the feedback unit 152 may monitor grid stability, temperature, power consumption of the serversP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0150] - 29 -
[0151] 121, 122 or may detect a malfunction of a component of the computer system 100.
[0152] The feedback unit 152 generates a feedback signal 4 if, for example, the sensor observes that a threshold is exceeded or a malfunction occurs . In this case the decision unit 153 as part of the infrastructure control unit 151 may decide to adapt or terminate the boot protocol 161, 162, 163 in order to reduce the risk of damaging the computer system 100.
[0153] In a first step SI of the method for booting the computer system 100 according to an exemplary embodiment, as illustrated in Figure 2, the uninterruptible power supply 180 and the computational unit 111 restart . Since the infrastructure control unit 151 is integrated in the uninterruptible power supply 180, the infrastructure control unit 151 essentially boots sequentially to the uninterruptible power supply 180. This means that the infrastructure control unit 151 and the power supply system 110 boot or restart prior to the servers 121, 122.
[0154] In the present exemplary embodiment, the power supply system 110 was shut down prior to carrying out the method, for example due to a power shortage . That is, in the present exemplary embodiment the trigger event is restorage of power .
[0155] In a further step S2, the power distribution unit 112 is supplied with power . In substep S2a the infrastructure control unit 151 looks for availability of the power distribution unit 112 via the first data connection 5. In a subsequent substep S2b, the power distribution unit 112 communicates availability.P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0156] - 30 -
[0157] In subsequent steps S3, S4, the power outlets 114, 115 are switched on, supplying the servers 121, 122 with power and booting the servers 121, 122.
[0158] In substeps S3a, S4a, the infrastructure control unit 151 checks for availability of the first and second severs 121, 122 via the second and third data connections 6, 7, respectively. In subsequent substeps S3b, S4b, the servers 121, 122 communicate availability.
[0159] In a subsequent step S5, the infrastructure control unit 150 connects to the first server 121 through SSH and launches a command for reconfiguration of the first server 121.
[0160] In a further step S6, the infrastructure control unit 151 connects to the second server 122 through web service and sets the configuration with a REST API call .
[0161] In substeps S5a, S6a, the infrastructure control unit 151 waits for the servers 121, 122 to be configured. In subsequent substeps S5b, S6b, the servers 121, 122 communicate termination of configuration and / or synchronization. In particular, the first server 121 and the second server 122 each provide an expected status .
[0162] In subsequent steps S7, S8 last initiation commands are executed on the servers 121, 122 to finalize the infrastructure boot or restart .
[0163] Advantageously, the method described in Figure 2 can be carried out fully automatically. Thus, no external interference, for example by a human technician or the like,P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0164] - 31 -
[0165] is necessary. Moreover, external devices such as relays for monitoring and coordinating the restart can be omitted.
[0166] In the present exemplary embodiment, all steps SI to S8 are controlled by the infrastructure control unit 151. In alternative examples it is however also possible that some of the steps are controlled by other components of the computer system 100. For example, the first server 121 may instruct configuration of the second server 122. However, at least mandatory start instructions of the complex server infrastructure are provided and controlled by the infrastructure control unit .
[0167] The computer system 100 according to the second exemplary embodiment as illustrated in Figure 3 comprises essentially the same features as the computer system 100 according to the first exemplary embodiment . In addition, the computer system 100 of Figure 3 comprises three sensors 131, 132, 133.
[0168] A first sensor 131 is a power sensor configured to detect whether the external power supply of the computer system 100 deviates from a defined saf e-to-operate threshold. If this saf e-to-operate threshold is exceeded, the computer system 100 may be shut down. This may be the case in the event of a power shortage . If the power falls back under the threshold, for example if power is restored, the first sensor 131 provides a first sensor signal 1. The first sensor 131 is an internal sensor integrated in the uninterruptable power supply 180.
[0169] A second sensor 132 is an environmental sensor configured to detect whether an environmental parameter deviates from a defined saf e-to-operate threshold or range . The environmentalP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0170] - 32 -
[0171] parameter may be temperature, humidity or the like . If the environmental parameter falls below the threshold or inside the range, the second sensor 132 provides a second sensor signal 2. The second sensor 132 is an external sensor remote from power supply 180 and the computer system 100.
[0172] A third sensor 133 is a clock. For example, the computer system 100 may be fully or partially shut down during certain periods of time, for example during weekends . By the clock 133, shutting down and restarting the computer system 100 can be scheduled and automatically performed. In this case, a third sensor signal 3 may be provided. The third sensor 133 is implemented as a software, for example .
[0173] The infrastructure control unit 151 is further configured to execute a plurality of boot protocols 161, 162, 163 and further comprises the decision unit 153. The decision unit 153 is configured to decide which boot protocol 161, 162, 163 is executed, in particular depending on the sensor signal 1, 2, 3 and the feedback signal 4. This means that different restart scenarios may require different boot protocols 161, 162, 163.
[0174] Furthermore, the decision unit 153 is configured to adapt or terminate a boot protocol 161, 162, 163 in response to the feedback signal 4.
[0175] Furthermore, the infrastructure control unit 151 comprises actuators 154, via which a power controller 200 of the external power supply of the computer system 100 may be controlled. If the feedback unit 152 detects that a safe-to-operate power consumption is exceeded, the actuators 154 can control the power controller 200 to reduce power supplied toP2025, 0025 WO N / P25-1004W001 January 28, 2026
[0176] 33
[0177] the computer system 100 to reduce a risk of damaging the computer system 100.
[0178] The infrastructure control unit 151 further comprises a controller 155. The controller 155 is configured to communicate, instruct and monitor the power distribution unit 112 and the servers 121, 122, in particular via the data connections 5, 6, 7. Such a controller 155 may be present in all other exemplary embodiments without restricting other features of these exemplary embodiments .
[0179] The invention is not restricted to the exemplary embodiments by the description on the basis of said exemplary embodiments . Rather, the invention encompasses any new feature and also any combination of features, which in particular comprises any combination of features in the patent claims and any combination of features in the exemplary embodiments, even if this feature or this combination itself is not explicitly specified in the patent claims or exemplary embodiments .P2025, 0025 WO N / P25-1004W001 January 28, 2026
[0180] - 34 -
[0181] List of reference numerals
[0182] 1, 2, 3 first, second, third sensor signal feedback signal
[0183] 5, 6, 7 first, second, third data connection
[0184] power connection
[0185] 100 computer system
[0186] 110 power supply system
[0187] 111 computational unit of power supply system 112 power distribution unit
[0188] 113 computational unit of power distribution unit 114, 115 power outlet
[0189] 121, 122 first, second server
[0190] 131, 132, 133 first, second, third sensor
[0191] 141, 142 first, second server control unit
[0192] 151 infrastructure control unit
[0193] 152 feedback unit
[0194] 153 decision unit
[0195] 154 actuator
[0196] 155 controller
[0197] 161, 162, 163 boot protocol
[0198] 180 uninterruptible power supply
[0199] 191 master load segment
[0200] 192 first load segment
[0201] 200 power controller
[0202] SI to S7 method steps
Claims
92025 , 0025 WO N / P25- 1004W001 January 28 , 2026Claims1 . Method for booting or restarting a computer system ( 100 ) comprising at least one server ( 121 , 122 ) , a power supply system ( 110 ) configured to supply power to the server ( 121 , 122 ) , and an infrastructure control unit ( 151 ) integrated in the power supply system ( 110 ) , comprising the steps of :- detecting of a trigger event ,- booting the server ( 121 , 122 ) by executing a boot protocol ( 161 , 162 , 163 ) , wherein at least some instructions of the boot protocol ( 161 , 162 , 163 ) are provided by the infrastructure control unit ( 151 ) ,- monitoring execution of the boot protocol ( 161 , 162 , 163 ) by the infrastructure control unit ( 151 ) ,wherein the instructions of the boot protocol ( 161 , 162 , 163 ) provided by the infrastructure control unit ( 151 ) comprise at least mandatory start instructions for booting the server ( 121 , 122 ) .2 . Method according to claim 1 , whereinthe power supply system ( 110 ) and the infrastructure control unit ( 151 ) are booted prior to the server ( 121 , 122 ) .3 . Method according to one of the preceding claims , wherein the boot protocol ( 161 , 162 , 163 ) includes steps of starting, configuring and synchroni zing the at least one server ( 121 , 122 ) .4 . Method according to one of the preceding claims , wherein the boot protocol ( 161 , 162 , 163 ) includes steps of controlling a power controller ( 155 ) of an uninterruptible power supply ( 180 ) included in the power supply system ( 110 ) to turn on or of f at least one load segment ( 191 , 192 ) toP2025, 0025 WO N / P25-1004W001 January 28, 2026- 36 -deliver electricity to other components of the computer system ( 100) .
5. Method according to one of the preceding claims, wherein the boot protocol ( 161, 162, 163) includes steps of connecting to a computational unit ( 113) of a power distribution unit ( 112 ) included in the power supply system ( 110) to control a power controller of the power distribution unit ( 112 ) to turn on or off at least one power outlet ( 114, 115) of the power distribution units ( 112 ) to provide electricity to other components of the computer system ( 100) .
6. Method according to one of the preceding claims, wherein the infrastructure control unit ( 151 ) is configured to execute a plurality of boot protocols ( 161, 162, 163) , and, depending on the trigger event, a corresponding boot protocol ( 161, 162, 163) is executed by the infrastructure control unit ( 151 ) .
7. Method according to one of the preceding claims, wherein the trigger event comprises one or more of the following events :- exceeding or falling below a threshold of an environmental parameter,- reaching a scheduled event for booting or restarting the computer system ( 100) ,- presence or absence of power externally provided to the power supply system ( 110) ,- exceeding or falling below a threshold of a quality parameter of the power externally provided to the power supply system ( 110) ,- presence or absence of an internal failure of a component of the computer system ( 100) .P2025 , 0025 WO N / P25- 1004W001 January 28 , 20268 . Method according to one of the preceding claims , wherein the computer system ( 100 ) comprises or is connected to at least one sensor ( 131 , 132 , 133 ) , and the sensor ( 131 , 132 , 133 ) detects the trigger event .9 . Method according to claim 8 , wherein the power supply system ( 110 ) comprises a decision unit ( 153 ) , and the decision unit ( 153 ) monitors at least one sensor signal ( 1 , 2 , 3 ) provided by the at least one sensor ( 131 , 132 , 133 ) and determines the presence of a trigger event depending on the monitored sensor signal ( 1 , 2 , 3 ) .10 . Method according to one of the preceding claims , wherein the infrastructure control unit ( 151 ) comprises a feedback unit ( 152 ) ,the feedback unit ( 152 ) provides a feedback signal ( 4 ) , and the infrastructure control unit ( 151 ) adapts or terminates the boot protocol ( 161 , 162 , 163 ) depending on the feedback signal ( 4 ) .11 . Method according to claim 10 , wherein the infrastructure control unit ( 151 ) controls power supplied to the power supply system ( 110 ) depending on the feedback signal ( 4 ) . 12 . Method according to one of the preceding claims , wherein the boot protocol ( 161 , 162 , 163 ) is executed by a computational unit ( 111 ) of the infrastructure control unit ( 151 ) or of the power supply system ( 110 ) .13 . Method according to one of the preceding claims , wherein the computer system ( 100 ) comprises a plurality of servers ( 121 , 122 ) and some of the servers ( 121 , 122 ) are prioriti zed in the boot protocol ( 161 , 162 , 163 ) .P2025 , 0025 WO N / P25- 1004W001 January 28 , 202614 . Method according to one of the preceding claims , wherein the infrastructure control unit ( 151 ) is connected to a server control unit ( 141 , 142 ) integrated in the at least one server ( 121 , 122 ) and the infrastructure control unit ( 151 ) instructs the server ( 121 , 122 ) to execute applications , to execute configuration of server systems and applications , and / or to start nested virtual computer systems .15 . Method according to claim 14 in connection with claim 10 or 1 , wherein the feedback unit ( 152 ) receives- signal , states and metrics of the at least one server ( 121 , 122 ) from the server control unit ( 141 , 142 ) , and / or- signal , state and metrics from power devices ( 112 , 180 ) of the power supply system ( 110 ) .16 . Computer system ( 100 ) comprising- at least one server ( 121 , 122 ) ,- a power supply system ( 110 ) configured to supply power to the server ( 121 , 122 ) , and- an infrastructure control unit ( 151 ) integrated in the power supply system ( 110 ) ,wherein- the infrastructure control unit ( 151 ) is configured to boot the servers ( 121 , 122 ) according to a predefined boot protocol ( 161 , 162 , 163 ) upon detection of a trigger event by providing at least some instructions of the boot protocol ( 161 , 162 , 163 ) , and- the infrastructure control unit ( 151 ) is configured to monitor execution of the boot protocol ( 161 , 162 , 163 ) .17 . Computer system ( 100 ) according to claim 16 , further comprising a sensor ( 131 , 132 , 133 ) , whereinP2025 , 0025 WO N / P25- 1004W001 January 28 , 2026- the sensor ( 131 , 132 , 133 ) is configured to provide a sensor signal ( 1 , 2 , 3 ) in response to the trigger event , - the infrastructure control unit ( 151 ) comprises a decision unit ( 153 ) configured to determine the presence of a trigger event depending on the sensor signal ( 1 , 2 , 3 ) , and- the sensor ( 131 , 132 , 133 ) comprises at least one of a clock, an environmental sensor, a power sensor .18 . Computer system ( 100 ) according to claim 16 or 17 , wherein- the power supply system ( 110 ) comprises an uninterruptible power supply ( 180 ) and at least one power distribution unit ( 112 ) electrically connected to the uninterruptible power supply ( 180 ) ,- each of the uninterruptible power supply ( 180 ) and the power distribution unit ( 112 ) comprises a computational unit ( 111 , 113 ) connected by a data connection ( 5 ) ,- one of the computational units ( 111 , 113 ) comprises the infrastructure control unit ( 151 ) .19 . Computer system ( 100 ) according to one of claims 16 to 18 , wherein- the power distribution unit ( 112 ) comprises at least one power outlet ( 114 , 115 ) , wherein each power outlet ( 114 , 115 ) is connected to one server ( 121 , 122 ) ,- the server comprises a server control unit ( 141 , 142 ) , and - the infrastructure control unit ( 151 ) is connected to the server control unit ( 141 , 142 ) by a data connection ( 6 , 7 ) .