TW201324115A - Computer system and boot managing method of computer system - Google Patents

Abstract

A computer system and a boot managing method of the computer system suited for a remote server are provided. The boot managing method includes the following steps: When a computer device managed by the remote server is in a shutdown state, a boot command is sent to the computer device through an Internet, and a counter is reset to countdown a time-out. When a Basic Input Output System (BIOS) start-OK log message is not received and the time-out is countdown completed, a reboot command is sent to the computer device through the Internet, and the counter is reset again to retest whether the computer device is boot completed. Thus, the remote server proceed a boot management with the computer device by using a cloud computing mechanism, so that the computer devices can be equipped with low-efficiency baseboard management controller (BMC) chips to reduce costs.

Description

Computer system and computer system boot management method

The present invention relates to a Baseboard Management Controller (BMC) technology, and more particularly to a computer system and a remote server for booting a plurality of computer devices.

Baseboard Management Controller (BMC) is the core processing controller in the Intelligent Platform Management Interface (IPMI) architecture, which can sense, monitor and record various working conditions in the server. The abnormal working condition is detected to execute the corresponding program to eliminate the abnormality in an instant. In addition, BMC has remote management, system status detection and recording, data tracking and system response, enabling operators to effectively manage a large number of servers while reducing maintenance costs.

In view of the power of BMC, BMC is becoming more and more important in the management of servers. While the performance and status of BMC in the server is gradually increasing, the cost of BMC chips is gradually becoming more expensive, especially for BMC chips with high computing performance.

With the rapid development of cloud computing technology, many research institutions and related application services gradually transfer complex computing and huge database from the local end to the cloud server connected by many servers, so that the network is smooth. In this case, the speed of the calculation and the reaction rate of the application service can be increased. In the current cloud server, each server also has a BMC for monitoring management. Therefore, whether the computing mechanism in the BMC and the cloud computing mechanism can be combined to enable the server to achieve high computing performance by using a low-cost BMC chip is an important direction that researchers can study.

The invention provides a booting management method for a computer system, so that the remote server can use the cloud computing mechanism to perform a power-on management process for the computer device, and the computer device uses a low-performance substrate management controller to reduce the cost.

In addition, the present invention provides a computer system in which a remote server in a computer system can perform a power-on management process on a computer device managed by the cloud computing mechanism, so that the computer device can be reduced by using a low-performance substrate management controller. cost.

The invention provides a booting management method for a computer system, which is suitable for use in a remote server in a computer system. The booting method of the computer system includes the following steps. When the computer device managed by the remote server is in the off state, a power-on command can be transmitted to the computer device via the network, and a counter is reset to start the countdown pause time. Before the countdown of the pause time is completed, it is checked whether the basic input/output system (BIOS) boot log information transmitted by the computer device is received via the network. And, when the BIOS startup log information is not received, and the pause time has been counted down, a reboot command is sent to the computer device via the network, and the counter is reset again to restart the countdown. The timeout.

In an embodiment of the present invention, the foregoing power-on management method further includes the following steps. When transmitting the above power-on command, the remote server resets a restart flag to zero. When transmitting the above reboot command, the remote server increments the above restart flag by one. When the restart flag is equal to a preset number, the remote server stops transmitting the above reboot command and stops resetting the counter, and performs an alert operation to inform the maintenance personnel of the computer system.

In an embodiment of the present invention, the foregoing power-on management method further includes the following steps. After the above pause time is counted down, and it is checked whether the above BIOS startup log information is received, the remote server should be able to check whether a BIOS start execution log message transmitted by the computer device is received.

In an embodiment of the present invention, the foregoing power-on management method further includes the following steps. After the above pause time is counted down, and it is checked whether the above BIOS startup log information is received, the remote server should be able to check whether a power boot log message transmitted by the computer device is received.

In an embodiment of the invention, the computer device includes a baseboard management controller (BMC) that can receive the boot command to perform a boot operation. In addition, the BMC will also process each operation of the computer device as a System Event Log (SEL), and trigger a cloud transmission event to transmit each system event log to the remote server via the network. Device. In addition, the above system event log may include the above-mentioned BIOS start execution log information, BIOS startup completion log information, and power startup log information.

Viewed from another aspect, the present invention provides a computer system including at least one computer device and a remote server. Each computer device includes a BMC, which can receive a boot command to perform a boot operation, and process each operation on the computer device as a system event log, and trigger a cloud transmission event to pass the system event log through the network. The path is transmitted to the remote server in a timely manner.

In the above, when the computer device is in the off state, the remote server can transmit a power-on command to the computer device via the network, and reset a counter to start the countdown pause time. Before the above pause time is counted down, the remote server will check whether a BIOS startup log message transmitted by the computer device is received via the network. If the BIOS startup log message has not been received after the pause time has been counted down, the remote server transmits a reboot command to the computer device via the network, and resets the counter to restart the countdown pause. time.

Following the above, please refer to the above description for the remaining implementation details of this computer system, and I will not repeat them here.

Based on the above, in order to make the computer device use the BMC with low computing performance, the BMC of the embodiment of the present invention continuously provides the system management log of the computer device to the remote server through the network, so that the BMC is powered on. The management mechanism is processed by the remote server in the cloud computing mechanism, and the BMC itself has no judgment function. Therefore, the remote server can use the time-out reciprocal mechanism (or the commonly known watch dog mechanism) to determine whether the BIOS of the computer device has been started after sending the power-on command. Complete the error recovery boot (FRB) mechanism for the computer device.

The above described features and advantages of the present invention will be more apparent from the following description.

DETAILED DESCRIPTION OF THE INVENTION Reference will now be made in detail to the exemplary embodiments embodiments In addition, wherever possible, the elements and/

As shown in FIG. 1, FIG. 1 is a schematic diagram of a computer system 100 in accordance with an embodiment of the present invention. The computer system 100 includes at least one computer device 110, a network 130, and a remote server 140, and each of the computer devices 110 also includes a substrate management controller (BMC) 120. The computer device 110 is exemplified herein by a plurality of servers, but the present invention is not limited thereto. In other embodiments, a personal computer with BMC 120 can be utilized as an implementation of computer device 110.

Therefore, the functional architectures of the substrate management controller 120 and the remote server 140 in accordance with the present embodiment are as shown in FIG. 2 and FIG. 3, and FIG. 2 and FIG. 3 are functions of the computer device 110 and the remote server 140, respectively. Block diagram. As shown in FIG. 2 and FIG. 3, in order to enable the BMC 120 with low computing performance, the BMC 120 of the embodiment of the present invention does not have a judgment function, and the BMC judgment mechanism utilizes the remote end in the cloud computing mechanism. The server processes it.

In detail, the BMC 120 of the present embodiment is different from the conventional BMC in that the BMC 120 does not include a Message Handler 50 and a Platform Event Filter (PEF) owned by the conventional BMC. 260. The Fault Resilient Booting (FRB) module 280 further transfers the complex judgment mechanism in the alert processing module 270 to the remote server 140, so that the judgment mechanism of the BMC 120 in the past utilizes the remote servo. The device 140 processes in the manner of a cloud computing mechanism. The FRB module 280 includes an error recovery counter (that is, the counter 290 in FIG. 3), which is described in detail in the power-on management method described below.

As shown in FIG. 2, the computer device 110 includes a BMC 120, and the BMC 120 itself retains only a plurality of sensors 210, an event receiver 215, a system event log (SEL) device 220, and a cloud transmission module. 230 and network card interface 240. Different types and uses of the sensor 210 will be installed throughout the computer device 110. The BMC 120 uses the sensor 210 or other sensing devices connected to other interfaces to detect the operation of the computer device, such as the temperature of the microprocessor, the fan speed, etc., and transmits all of the operations to the The event receiver 215 is organized into a system event log (SEL).

The system event loger 220 records these system event logs. Then, the BMC 120 immediately triggers the cloud transmission event in the cloud transmission module 230 after the event receiver 215 generates the system event log, so that the system event log is transmitted through the network 130 by using the cloud transmission module 230 and the network card interface 240. Passed to the remote server 140 of FIG. In addition, the BMC 120 can also use the network card interface 240 and the cloud transmission module 230 to receive and execute program commands transmitted by the remote server 140 of FIG. 1 to the computer device 110.

In other embodiments, if it is desired to save the manufacturing cost of the BMC 120, the system event loger 220 can also be disposed in the remote server 140 of FIG. 3 instead of being disposed in the BMC 120, so that the BMC 120 is also There is no need to have a storage function with a system event log.

Referring to FIG. 3, the network card interface 240 and the cloud transmission module 230 of the remote server 140 can receive the system event log transmitted by the computer device 110 from the network 130, and pass the BMC type judgment process and record. After the source network address of the BMC, the signal processor 250 cooperates with the platform event filter 260 to analyze the system event log to determine whether the computer device is operating normally.

When it is determined that the system is abnormal (for example, the microprocessor is overheated, the fan is not operating normally, etc.), the signal processor 250 and the platform event filter 260 generate corresponding program commands, and through the network card interface 240 and the cloud. The transfer module 230 passes the program commands to the BMC 120 of the corresponding computer device 110. The cloud transmission event of the cloud transmission module 230 is implemented by using a software, but can also be implemented by using a modular method in hardware, thereby reducing development cost, and thus the present invention is not limited thereto.

The remote server 140 can also use the alert processing module 260 to perform an alert operation to notify the maintenance personnel to process the computer device 110. On the other hand, the remote server 140 can also use the error recovery boot module 280 to perform the boot management/boot reply procedure of the computer device 110 under its management.

In particular, in the past, when each BMC was in the boot management process, the conventional practice was to use the Fault Resilient Booting (FRB) module in the BMC to determine the basic input/output system (Basic Input/Output). System; BIOS) Whether to complete or not in a time-out, in order to achieve the boot management / error recovery boot mechanism. However, if the computing mechanism of the BMC 120 is implemented by the cloud computing mechanism, the BMC 120 does not have the above FRB module, and thus the computer device 110 cannot be properly booted.

Therefore, the spirit of the present invention is to use the cloud computing mechanism to move the boot management/error recovery boot mechanism function of the BMC 120 in each computer device 110 to the remote server 140, thereby reducing the operational efficiency of the baseboard management controller. , thereby reducing the cost of building the BMC 120.

Herein, an embodiment is proposed herein to achieve the spirit of the present invention. FIG. 4 is a schematic diagram of a booting method of a computer system 100 according to an embodiment of the invention, which is applicable to a remote server 140 in the computer system 100. In addition, the BMC 120 of the computer device 110 also needs to be correspondingly designed corresponding to the boot management method.

4 and FIG. 3, in the booting management method of the computer system 100, a time-out is preset in the counter 290. If the remote server 140 knows that one of the computer devices 110 is in a power-off state in all system event logs of the computer device 110 under its management, and the computer device 110 should be powered on; or if When the shutdown computer device 110 is turned on, the process proceeds to step S410. The remote server 140 transmits a power-on command to the computer device 110 via the network 130 through the cloud transmission module 230 and the network card interface 240.

Moreover, in step S410 of the embodiment, the counter 290 in the error recovery power-on module 280 will be reset and start counting down the preset pause time. In addition, the error recovery boot module 280 also resets a restart flag Frst to zero. In other embodiments, the counter 290 can also be implemented by using a watchdog timer (Watch Dog Timer), and details are not described herein again.

In step S420, the error recovery boot module 280 of the remote server 140 continuously determines whether the pause time is counted down. Before the timeout of the pause time is completed, the remote server 140 will continuously check whether the basic input/output system (BIOS) startup completion log information transmitted by the computer device 110 is received via the network 130 (step S430). In this embodiment, the BIOS startup log information referred to herein is a type of system event log (SEL), and the BIOS in the computer device 110 initiates a power-on self-test (POST). A system event log that will be generated when the program execution is complete.

Normally, if the remote server 140 receives the BIOS startup log information before the pause time is counted down, it indicates that the computer device 110 has been powered on, so the process proceeds to step S440, and the remote server 140 is at the computer device. The state of 110 is noted as being powered on.

However, unfortunately, if the BIOS startup log information of the computer device 110 is not received, and the pause time in step S420 has been counted down, the process proceeds from step S420 to step S450, and the error recovery boot module 280 checks the restart. Whether the startup flag Frst exceeds a predetermined number of restarts, so that the remote server 140 does not have to continuously restart the computer device 110. In the present embodiment, the predetermined number described above may be 5, but is not limited thereto.

If the restart flag Frst does not equal or exceed the predetermined number, the process proceeds from step S450 to step S460, and the remote server 140 transmits a reboot command via the network 130 or transmits a power-on command to the corresponding one. The computer device 110, the error recovery boot module 280 also resets the counter 290 again to return to step S420, and restarts the countdown pause time. In addition, the error recovery boot module 280 simultaneously adds 1 to the restart flag Frst to count the number of times the remote server 140 reboots the computer device 110.

In contrast, if the restart flag Frst has equaled or exceeded the predetermined number, it indicates that the number of times the computer device 110 has been rebooted has reached a predetermined number of times. Therefore, proceeding from step S450 to step S470, the remote server 140 stops transmitting the above-mentioned restart command, and the error recovery boot module 280 also stops resetting the counter 290.

Moreover, in step S470, the remote server 140 uses the alert processing module 270 to perform an alert operation to automatically inform the maintenance personnel of the computer system 100 of the information and request assistance from the maintenance personnel. The above warning operation may be: sending specific warning information on the screen of the remote server 140, transmitting a specific warning packet to a specific server or issuing a specific warning sound, etc., and the present invention should not be limited to the above examples.

In the above description, when the BMC 120 in each of the computer devices 110 deletes the relevant module of the boot management/error recovery booting mechanism, the remote server 140 of the embodiment of the present invention can transmit the error recovery boot mode. Group 280 achieves the same effect as previous BMC for each computer device 110. Compared with the conventional computer system, the embodiment of the present invention can utilize the BMC with low computational efficiency to achieve the related mechanism of the BMC with high computational efficiency, so that the installation cost of each computer device 110 can be further reduced. Specifically, as long as the counter 290 in the error recovery boot module 280 has a sufficient number, the remote server 140 can simultaneously perform the above boot management method/error for a plurality of computer devices 110 managed by the remote server 140. Reply to the boot mechanism.

FIG. 5 is a diagram showing a booting management method of a computer system 100 according to another embodiment of the present invention. This embodiment is similar to the power-on management method shown in FIG. 4 above, and the same or similar descriptions are not described herein again. The difference between this embodiment and the embodiment of FIG. 4 is that, in addition to determining the BIOS startup log information, the previous BMC needs to be related to other error recovery startup mechanisms in the boot management method/error recovery startup mechanism. The system event information is determined, for example, the power-on log information of the computer device 110 (the system event log generated when the computer device 110 is powered on), and the BIOS start executing the log information (the system generated when the BIOS starts executing in the computer device 110) Event log).

Therefore, in step S530 of FIG. 5, before the pause time countdown is completed, the remote server 140 will continuously check whether the power boot log information transmitted by the computer device 110 is sequentially received via the network 130 (ie, first. Verifying whether the computer device 110 is powered on), and then continuing to check whether the receiving BIOS starts executing log information (ie, checking whether the BIOS of the computer device 110 is started), and finally checking whether to receive the above BIOS startup log information (ie, verifying Whether the BIO of the computer device 110 is completed or not). If the system event logs are received in the order described above, then the process proceeds to step S440 to confirm that the computer device 110 has been powered on. In contrast, if the system event logs are not received in the above order, it indicates that the computer device 110 has generated an error in its booting procedure, so the process proceeds to steps S450 to S470 to continue the error recovery booting mechanism.

In summary, in the embodiment of the present invention, in order to enable the computer device to adopt a low-performance BMC, the BMC of the embodiment of the present invention continuously provides the system management log of the computer device to the remote server through the network, so that the BMC The boot management mechanism is processed by the remote server in the cloud computing mechanism, and the BMC itself has no judgment function. Therefore, the remote server can use the time-out reciprocal mechanism (or the commonly known watch dog mechanism) to determine whether the BIOS of the computer device has been started after sending the power-on command. Complete the boot management mechanism (or called the error recovery boot (FRB) mechanism) for the computer device.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the invention, and any one of ordinary skill in the art can make some modifications and refinements without departing from the spirit and scope of the invention. The scope of the invention is defined by the scope of the appended claims.

100. . . computer system

110. . . Computer device

120. . . Baseboard Management Controller (BMC)

130. . . network

140. . . Remote server

210. . . Sensor

215. . . Event receiver

220. . . System event logger

230. . . Cloud transmission module

240. . . Network card interface

250. . . Signal processor

260. . . Platform event filter

270. . . Warning processing module

280. . . Error recovery boot module

290. . . counter

S410~S530. . . step

1 is a schematic diagram of a computer system in accordance with an embodiment of the invention.

2 is a functional block diagram of a computer device.

Figure 3 is a functional block diagram of the remote server.

FIG. 4 is a diagram showing a method for booting up a computer system according to an embodiment of the invention.

FIG. 5 is a diagram showing a method for booting up a computer system according to another embodiment of the present invention.

S410~S530. . . step

Claims (10)

A booting management method for a computer system is applicable to a remote server, and the booting management method of the computer system includes: when a computer device managed by the remote server is in a shutdown state, transmitting a boot command via a network Go to the computer device and reset a counter in the remote server to start the countdown pause time; before the countdown is completed, check whether a BIOS transmitted by the computer device is received via the network. Log information; and when the BIOS startup log information is not received, and the pause time has been counted down, a reboot command is transmitted to the computer device via the network, and the counter is reset to restart the countdown Pause time. The booting management method of the computer system according to claim 1, further comprising: resetting a restart flag to 0 when transmitting the power-on command; and re-setting the restart command when transmitting the power-on command The startup flag is incremented by one; and when the restart flag is equal to a preset number, the transmission of the restart command is stopped and the resetting of the counter is stopped, and an alert operation is performed. The method for controlling the booting of the computer system according to claim 1, further comprising: checking whether the computer device transmits the received information after receiving the BIOS startup log information after the timeout is completed. The BIOS begins to execute log information. The method for controlling the booting of the computer system according to claim 1, further comprising: checking whether the computer device transmits the received information after receiving the BIOS startup log information after the timeout is completed. Power boot log information. The computer system boot management method according to claim 1, wherein the computer device comprises: a base management controller, which receives the boot command to perform a boot operation, and each operation of the computer device The situation is processed into at least one system event log and triggers a cloud transmission event to communicate the system event log to the remote server via the network. The computer system boot management method of claim 5, wherein the system event log includes a BIOS start execution log information and the BIOS startup log information. A computer system comprising: at least one computer device, comprising: a substrate management controller, receiving a power-on command to perform a power-on operation, and processing each operation condition of the computer device as at least one system event log, and triggering a cloud transmission event to transmit the system event log to the remote server via a network; and a remote server including a counter, wherein the computer device managed by the remote server is a In the shutdown state, the remote server transmits a power-on command to the computer device via the network, and resets the counter to start a countdown pause time; before the pause time is counted down, the remote server checks whether Receiving, by the network, a BIOS startup log information transmitted by the computer device; when the BIOS startup log information is not received, and the pause time has been counted down, the remote server transmits a re-transmission via the network. Power on the command to the computer device and reset the counter to restart the countdown. The computer system of claim 7, wherein the remote server resets a restart flag to 0 when transmitting the power-on command, and the remote server transmits the restart command when transmitting the power-on command The restart flag is further incremented by 1. When the restart flag is equal to a preset number, the remote server stops transmitting the restart command, stops resetting the counter, and performs an alert operation. The computer system of claim 8, wherein the remote server checks whether a BIOS transmitted by the computer device is received before the timeout is completed and the BIOS startup log information is received. Start executing log information. The computer system of claim 8, wherein the remote server checks whether a power source transmitted by the computer device is received before the timeout is completed and the BIOS startup log information is received. Start log information.