TW201715396A - Server and error detecting method thereof - Google Patents

Abstract

The invention provides an error detecting method for a server that includes a motherboard having a controller chipset and a baseboard management controller (BMC). The method comprises steps of: Storing, by the BMC, an error comparison table of relationships between system error codes and system error messages; Providing, by the controller chipset, at least one system error code when the server boots abnormally; Receiving, by the BMC, the system error code from the controller chipset; Matching, by the BMC, based on the error comparison table, the received system error code to at least one system error message; Detecting, by the BMC, whether there is a message output request from outside of the motherboard in order to determine whether or not to output the system error message.

Description

Server and its debugging method

The present invention relates to an electronic data processing apparatus and method, and more particularly to a server for detecting an abnormality in power-on detection and a method for detecting the same.

With the development of technologies such as the Internet of Things, the cloud, and big data, the requirements for the stability and reliability of the server are getting higher and higher. The server is the hardware foundation of high-performance computing technologies such as the cloud and the Internet of Things. The key to rapid development. The existing server uses a variety of components, such as microprocessors, memory, motherboards, hardware boards (such as display cards, sound cards, etc.), fans and other heat sinks, and with the development of server technology, various hardware Most of the boards, devices, etc. need to be integrated on the motherboard. Therefore, when the server fails to start and crash during the boot process, how to diagnose the problem through the debugging of the motherboard is an important factor to ensure the normal operation of the server. Question.

In general, the server's motherboard includes a south bridge and a basic input/output system (BIOS), and an 80-bit decoder (port 80 decoder) is connected to the debug detection during the boot process. And a seven-segment display, the low-pin number (Low Pin) of the south bridge during the boot process The Count) interface outputs a power-on detection signal, and the 80-inch decoder receives the power-on detection signal of the south bridge, and then decodes the octet signal to control the seven-segment display to display a corresponding error code, for example, displaying C6 to indicate dynamic random access. Memory (DRAM) failure.

However, the disadvantage of this method is that if the user needs to repair and correct the error, the server needs to be opened and the other devices are removed, and the 80-inch decoder and the seven-segment display are connected to observe the error code; and the user cannot see at a glance You can see the fault, you must judge by experience, or consult the manual to understand the relative meaning of the error code C6, and this is only the error code displayed corresponding to the last detection error. In an environment without a network, or even when the user does not have the 80-inch decoder and the seven-segment display at all, the user will lack other channels to get debugging information and is more helpless.

Accordingly, it is an object of the present invention to provide an error detection method suitable for use in a server, and the server includes a motherboard having a control chip set and a substrate management electrically connected to the control chip set The controller, and the debugging method comprises a step (A), a step (B), a step (C), a step (D), and a step (F).

The step (A) is that the baseboard management controller stores an error comparison table, and the error comparison table records a plurality of different system error codes and a plurality of different system error messages, each system error code corresponding to a system error message, and different The system error message is related to a variety of different server boot exceptions.

The step (B) is that when the server is powered on abnormally, the control chipset outputs at least one system error code.

The step (C) is that the substrate management controller receives the at least one system error code from the control chip set.

The step (D) is that the baseboard management controller finds at least one corresponding system error message from the error comparison table according to the at least one system error code.

The step (F) is that the baseboard management controller determines whether to receive a message output request from outside the motherboard to determine whether to output the at least one system error message.

The effect of the present invention is that the substrate management controller finds the corresponding system error message from the stored error comparison table according to the received system error code, and outputs the corresponding system error message when receiving the message output request. The system error message allows the user to quickly and directly know what the error is.

1‧‧‧Server

11‧‧‧ motherboard

110‧‧‧Error comparison table

111‧‧‧Control chipset

1111‧‧‧Detection error

113‧‧‧Base management controller

1131‧‧‧ Non-volatile memory

1132‧‧‧RS232 buffer

1133‧‧‧Video Graphics Conversion Module

1134‧‧‧ memory card

1135‧‧‧Remote serial communication module

114‧‧‧Basic input and output unit

115‧‧‧Video Graphics Conversion Interface

116‧‧‧RS232 busbar

12‧‧‧Storage device

121‧‧‧Message Acquisition Module

13‧‧‧ display

14‧‧‧Electronic devices

141‧‧‧Message Acquisition Module

142‧‧‧Internet User Interface

15‧‧‧Electronic devices

151‧‧‧Display module

152‧‧‧ Remote serial communication interface

153‧‧‧Console redirection module

A~I‧‧‧ steps

F1~F8‧‧‧ substeps

Other features and advantages of the present invention will be apparent from the embodiments of the present invention, wherein: FIG. 1 is a system block diagram illustrating an embodiment of a server of the present invention; FIG. 2 is a schematic diagram illustrating the present invention. a substrate management controller of the server and a fault comparison table stored by a basic input/output unit; FIG. 3 is a system block diagram illustrating a first embodiment of the detection method of the present invention; 4 is a flow chart, the first embodiment is illustrated in FIG. 3; FIG. 5 is a system block diagram showing a second embodiment of the detecting method of the present invention; FIG. 6 is a flowchart, and FIG. FIG. 7 is a system block diagram showing a third embodiment of the detecting method of the present invention; and FIG. 8 is a flowchart, the auxiliary FIG. 7 illustrates the third embodiment; FIG. 9 is a system block. The figure illustrates a fourth embodiment of the detecting method of the present invention; and FIG. 10 is a flowchart, and FIG. 9 illustrates the fourth embodiment.

Before the present invention is described in detail, it should be noted that in the following description, similar elements are denoted by the same reference numerals.

Referring to FIG. 1 , an embodiment of the server 1 of the present invention includes a motherboard 11 , a storage device 12 , a display 13 , an electronic device 14 , and an electronic device 15 .

The motherboard 11 includes a control chipset 111, a baseboard management controller 113, a basic input/output system 114, a video graphics conversion interface 115, and an RS232 busbar 116. .

The control chip set 111 has an error detection 1111 for outputting at least one system error code when a power-on abnormality occurs during the power-on detection of the server 1. In this example, the control chip set 111 is at the platform controller center (Platform Controller The Hub 来 1111 is implemented, for example, the 80th (port 80), the 81st (port 81), and the like.

The substrate management controller 113 includes a non-volatile memory 1131, an RS232 buffer 1132, a video graphics conversion module 1133, and a serial-over-LAN communication module 1135. The baseboard management controller 113 stores an error comparison table 110, which records a plurality of different system error codes and a plurality of different system error messages, and the baseboard management controller 113 electrically connects the control chipset via the LPC busbars. Detecting 1111 of 111 to receive at least one system error code from the control chip set 111, and thereby finding the corresponding at least one system error message from the error comparison table 110, recording the non-volatile memory in the non-volatile memory 1131. For details of the operation of the baseboard management controller 113, for example, the RS232 buffer 1132, the video graphics conversion module 1133, and the execution details of the remote serial communication module 1135, please refer to it later. The non-volatile memory 1131 used by the substrate management controller 113 to record the system error message may be the non-volatile random access memory (NVRAM) of the substrate management controller 113 itself. The baseboard management controller 113 can also record the system error message on an external memory card 1134 such as a Secure Digital Memory Card, both of which can store a plurality of system error messages. In addition, the video graphics conversion module 1133 is, for example, a video graphics array (VGA) module, and is electrically connected to the control chip group 111 via a fast peripheral interconnect standard (PCIE) bus, and the video graphics conversion module 1133 is used for input. And outputting, by the substrate management controller 113, the at least one display signal obtained by converting the at least one system error message.

Referring to FIG. 2, in the embodiment, the error comparison table 110 defines the system error code and the system error message when the server 1 fails according to an industry standard issued by, for example, an AMI basic input/output unit manufacturer. Correspondence relationship, and in the error comparison table 110, each system error code corresponds to a system error message, and various system error messages are related to various server startup abnormal conditions, for example: system error codes 0xA1, 0xA2, 0xA3 0xA4, 0xA5, 0xA6, 0xA7, etc. correspond to memory not installed, CPU type or speed is not supported, CPU mismatch, CPU mismatch, CPU initialization error, North bridge initialization error, South bridge initialization error, Flash update is failed, etc.

Referring to FIG. 1, the basic input output unit 114 stores the error comparison table 110, and the basic input output unit 114 is electrically connected to the debug layer 1111 of the control chip group 111 via a serial peripheral interface (SPI) bus to receive the error. The system error code is used to find the corresponding at least one system error message from the error comparison table 110. Moreover, the basic input/output unit 114 can convert the at least one system error message into At least one of the signals is displayed and output. In the present embodiment, the error comparison table 110 stored by the basic input/output unit 114 is the same as the error comparison table 110 stored by the substrate management controller 113.

The video graphics conversion interface 115 is electrically connected to the video graphics conversion module 1133 of the substrate management controller 113, and can receive the display signal output from the video graphics conversion module 1133 of the substrate management controller 113, or receive the received signal from The basic input/output unit 114 outputs the display signal transmitted via the control chip group 111 and outputs it.

The RS232 bus bar 116 is electrically coupled to the RS232 buffer 1132 of the baseboard management controller 113, the details of which are described in detail below.

The storage device 12 is electrically connected to the control chip set 111 via a USB/SATA bus, and the storage device 12 has a message acquisition module 121 capable of issuing a message output request to acquire the substrate management via the control chip set 111. The at least one system error message output by the controller 113. The storage device 12 can be implemented using a flash drive, a SATA hard disk, or a solid state drive (SSD).

The electronic device 14 is connected to the baseboard management controller 113 via a network, and the electronic device 14 has a message acquisition module 141 and a web UI 142. The message obtaining module 141 of the electronic device 14 can send the message output request to obtain the at least one system output by the baseboard management controller 113 through the network. Error message. The network user interface 142 is used to display the system error message. In this example, the electronic device 14 is, for example, a desktop computer or a notebook computer.

The display 13 is electrically connected to the video graphics conversion interface 115 of the motherboard 11 to receive the display signal for display.

The electronic device 15 has a display module 151, a remote serial communication interface 152 capable of communicating with the remote serial communication module 1135 of the baseboard management controller 113, and an electrical connection of the motherboard 11. The console redirection utility 153 of the RS232 bus bar 116. In this example, the electronic device 15 is, for example, a desktop computer or a notebook computer.

Referring to Figures 3 and 4, a first embodiment of the debug method of the present invention is implemented in the server 1 shown in Figure 3. The respective steps of the first embodiment will be described in detail below.

In step A, the substrate management controller 113 stores the error comparison table 110.

In step B, when the server 1 is turned on abnormally, the control chip set 111 outputs at least one system error code.

In step C, the baseboard management controller 113 receives the at least one system error code from the control chipset 111.

In step D, the baseboard management controller 113 finds the corresponding at least one system error message from the error comparison table 110 according to the at least one system error code.

In step E, the substrate management controller 113 records the at least one system error message in the non-volatile memory 1131. Further, the baseboard management controller 113 records a plurality of system error messages in the non-volatile memory 1131, and/or records a plurality of system error messages on the memory card 1134 to ensure that the server 1 is powered down or when At the time of the machine, all system errors that have occurred can be retained for the user to know.

In step F, the baseboard management controller 113 determines whether a message output request from the outside of the motherboard 11 is received to determine whether to output the at least one system error message. In detail, this step F includes the following sub-steps.

In step F1, when the message obtaining module 121 of the storage device 12 issues the message output request by the local user operation, the message output request is transmitted to the baseboard management controller 113 via the control chip group 111. It is explained here that the message output request is an Intelligent Platform Management Interface (IPMI) instruction.

In step F2, when the substrate management controller 113 receives the message output request from the storage device 12, the at least one system error message is outputted, and is stored in the storage device 12 via the control chipset 111, thereby being locally available. The end user quickly and directly knows all the system errors that the server 1 has encountered during the power-on detection process. In addition, the substrate management controller 113 outputs all system error messages recorded in the non-volatile memory 1131 (or the memory card 1134), so that the local end user can completely know that all the occurrences have occurred. System error condition.

Referring to Figures 5 and 6, a second embodiment of the debug method of the present invention is implemented in the server 1 shown in Figure 5. This second embodiment is the same as step A to step E of the first embodiment, with the difference being the step F, and the step G to the step J after the step F. Wherein, the step F includes the following sub-steps.

In step F3, when the display 13 is connected to the video graphics conversion interface 115 of the motherboard 11, the user sends a video graphics conversion signal as the message output request to the video graphics conversion module of the baseboard management controller 113. Group 1133.

In step F4, when the video graphics conversion module 1133 of the baseboard management controller 113 receives the message output request from the display 13 via the video graphics conversion interface 115, the video graphics conversion module 1133 of the baseboard management controller 113 The at least one system error message is converted into at least one display signal, and is output to the display 13 via the video graphics conversion interface 115 of the motherboard 11.

In step G, the basic input/output unit 114 initializes the video graphics conversion module 1133 of the baseboard management controller 113, and after the initialization is completed, notifies the baseboard management controller 113 to stop the output of the baseboard management controller 113. The display signal is taken over by the basic input/output unit 114 to the video graphics conversion module 1133.

In step H, the basic input output unit 114 receives the at least one system error code from the control chip set 111, and accordingly finds the corresponding at least one system error message from the error comparison table 110.

In step I, the basic input/output unit 114 converts the at least one system error message into at least one display signal, and outputs the video graphics conversion interface to the motherboard 11 via the video graphics conversion module 1133 of the baseboard management controller 113. 115 is further output to the display 13. In this way, even if there is no network or the storage device 12, the user can view the display 13 to know the system error that occurs during the boot detection process.

Referring to Figures 7 and 8, a third embodiment of the debug method of the present invention is implemented in the server 1 shown in Figure 7. This third embodiment is identical to steps A through E of the first embodiment, with the difference being in step F, and the step F includes the following sub-steps.

In step F5, when the message obtaining module 141 of the electronic device 14 issues the message output request by the remote user operation, the message output request is transmitted to the baseboard management controller 113 via the network. It is explained here that the message output request is an Intelligent Platform Management Interface (IPMI) instruction.

In step F6, when the substrate management controller 113 receives the message output request from the electronic device 14, the at least one system error message is output to the network user interface 142 of the electronic device 14 via the network. Thus, even in a high-density server room, when the number of servers 1 is large, the user can directly pass the electrons. The network user interface 142 of the device 14 quickly knows which server 1 of the server 1 has an error during the power-on detection process, and can directly recognize the error of the system error message.

Referring to Figures 9 and 10, a fourth embodiment of the debug method of the present invention is implemented in the server 1 shown in Figure 9. This fourth embodiment is identical to steps A through E of the first embodiment, with the difference being in step F, and the step F includes the following sub-steps.

In step F7, the electronic device 15 can send the message output request to the baseboard management controller by the user operation, wherein the message output request includes a console redirection command and a remote end. One of the serial-over-LAN communication instructions. In detail, the local end user can issue the console redirection command by operating the console redirection module 153 of the electronic device 15 via the RS232 bus bar 116 of the motherboard 11 to the baseboard management controller 113. The remote user can use the remote serial communication interface 152 of the electronic device 15 to send the remote serial communication command to the baseboard management controller 113 through the network.

In step F8, when the substrate management controller 113 converts the at least one system error message into at least one RS232 message, stores it in the RS232 buffer 1132, and upon receiving the message output request from the electronic device 15, the at least An RS232 message is output from the RS232 buffer 1132 to the electronic device 15 via the RS232 bus 116. In more detail, for local users, this The RS232 message can be output to the console redirection module 153 of the electronic device 15 via the RS232 bus bar 116. In addition, the RS232 message can also be output to the remote serial communication interface 152 of the electronic device 15 through the remote serial communication module 1135 of the baseboard management controller 113, and displayed and used by the display module 151. The person knows. In this way, regardless of whether the user's environment has a network, or in the server room or the remote end, it can quickly and directly know which server 1 of the server 1 has an error during the boot detection process, and can increase The flexibility of the user in repairing the fault.

In summary, with the server of the present invention and its debugging method, when the user does not have, for example, an 80-inch decoder or a seven-segment display to display debugging, the user can use the network according to the environment in which he is located. The storage device 12, the display 13, the electronic device 14, and the electronic device 15 communicate with the substrate management controller 113, not only can quickly and directly acquire specific system error messages, but also increase the pipeline and the user's debugging. Flexibility in different environmental applications. In addition, by recording the plurality of system error messages by the non-volatile memory 1131 or the memory card 1134, the user can completely know all the system error conditions that have occurred, and thus the object of the present invention can be achieved.

However, the above is only the embodiment of the present invention, and the scope of the invention is not limited thereto, and all the equivalent equivalent changes and modifications according to the scope of the patent application and the patent specification of the present invention are still The scope of the invention is covered.

1‧‧‧Server

11‧‧‧ motherboard

110‧‧‧Error comparison table

111‧‧‧Control chipset

1111‧‧‧Detection error

113‧‧‧Base management controller

1131‧‧‧ Non-volatile memory

1132‧‧‧RS232 buffer

1133‧‧‧Video Graphics Conversion Module

1134‧‧‧ memory card

1135‧‧‧Remote serial communication module

114‧‧‧Basic input and output unit

115‧‧‧Video Graphics Conversion Interface

116‧‧‧RS232 busbar

12‧‧‧Storage device

121‧‧‧Message Acquisition Module

13‧‧‧ display

14‧‧‧Electronic devices

141‧‧‧Message Acquisition Module

142‧‧‧Internet User Interface

15‧‧‧Electronic devices

151‧‧‧Display module

152‧‧‧ Remote serial communication interface

153‧‧‧Console redirection module

Claims (14)

An error detection method is applicable to a server, the server includes a motherboard, and the motherboard has a control chip set, and a substrate management controller electrically connected to the control chip set, the debugging method includes: A) the substrate management controller stores an error comparison table, and the error comparison table records a plurality of different system error codes and a plurality of different system error messages, each system error code corresponding to one system error message, and multiple different system error messages Respectively related to a plurality of different server boot abnormalities; (B) when the server is powered on abnormally, the control chipset outputs at least one system error code; (C) the baseboard management controller receives the control chipset The at least one system error code; (D) the baseboard management controller finds the corresponding at least one system error message from the error comparison table according to the at least one system error code; and (F) the baseboard management controller Determining whether to receive a message output request from outside the motherboard to determine whether to output the at least one system error message. The method of claim 1, wherein the baseboard management controller has a non-volatile memory, the debug method further comprising a step (E) between the step (D) and the step (F): E) the substrate management controller records the at least one system error message in the non-volatile memory, And in the step (F), if the substrate management controller receives the message output request, the at least one system error message in the non-volatile memory is output. According to the debugging method described in claim 1, the control chip group of the motherboard is electrically connected to a storage device, and in the step (F), when the substrate management controller receives the message output request from the storage device At this time, the at least one system error message is outputted and stored in the storage device via the control chip group. According to the debugging method described in claim 1, the baseboard management controller of the motherboard connects an electronic device through the network, and in the step (F), when the baseboard management controller receives the message output request from the electronic device At the same time, the at least one system error message is output to the electronic device via the network. According to the debugging method of claim 1, the baseboard management controller of the motherboard is connected to an electronic device, and in the step (F), when the baseboard management controller receives the message output request from the electronic device Translating the at least one system error message into at least one RS232 message for outputting to the electronic device, wherein the message output request includes a console redirection command and a remote-over-LAN One of the communication instructions. According to the debugging method of claim 1, the substrate management controller of the motherboard is electrically connected to a display, and in the step (F), when the substrate management controller receives a video graphics conversion signal from the display As the message output request, the baseboard management controller converts the at least one system error message into at least one display signal for output to the display. The debugging method of claim 6, the motherboard further has a video graphics conversion interface electrically connected to the display, and a basic input/output unit electrically connected to the control chipset and storing the error comparison table, the motherboard The baseboard management controller further includes a video graphics conversion module electrically connected to the video graphics conversion interface and the control chipset, and the debugging method further includes: (G) the basic input/output unit initializing the video graphics conversion module And after the initialization is completed, notifying the baseboard management controller to cause the baseboard management controller to stop outputting the display signal; (H) the basic input/output unit receives the at least one system error code from the control chipset, and Corresponding to the corresponding at least one system error message from the error comparison table; and (I) the basic input/output unit converts the at least one system error message into at least one display signal, and the video is managed by the baseboard management controller. The graphics conversion module and the video graphics conversion interface of the motherboard are outputted to the display. A server includes: a motherboard, including a control chipset, having a debug error to output at least one system error code when the server is powered on abnormally; and a substrate management controller electrically connecting the control chipset Debugging, and storing an error comparison table, which records a plurality of different system error codes and a plurality of different system error messages, each system error code corresponding to a system error message, and various system error messages are related to Various different server boot abnormal conditions, The baseboard management controller receives at least one system error code from the control chipset, and accordingly finds the corresponding at least one system error message from the error comparison table, and the baseboard management controller determines whether to receive from the motherboard An external message output request determines whether to output the at least one system error message. The server of claim 8, wherein the substrate management controller has a non-volatile memory, and the substrate management controller records the at least one system error message in the non-volatile memory, and wherein Receiving the message output request, the baseboard management controller outputs the at least one system error message in the non-volatile memory. The server of claim 8, wherein the control chipset of the motherboard is further electrically connected to a storage device, the storage device has a message acquisition module capable of issuing the message output request, and when the substrate management controller When receiving the message output request from the message acquisition module of the storage device, the at least one system error message is outputted and stored in the storage device via the control chip group. The server of claim 8, wherein the motherboard management controller of the motherboard is connected to an electronic device through a network, the electronic device has a message acquisition module capable of issuing the message output request, and when the substrate is managed When receiving the message output request from the message acquisition module of the electronic device, the controller outputs the at least one system error message to the electronic device via the network. The server of claim 8 is adapted to be connected to an electronic device, wherein the baseboard management controller of the motherboard has an RS232 buffer connected to the electronic device, and when the baseboard management controller receives the RS232 buffer When the message output request from the electronic device is requested, the at least one system error message is converted into at least one RS232 message, and is output to the electronic device via the RS232 buffer, wherein the message output request includes a console redirection (console redirection) An instruction and one of a serial-over-LAN communication instruction. The server of claim 8 is adapted to be electrically connected to a display, wherein the motherboard further includes a video graphics conversion interface electrically connected to the display, and the baseboard management controller of the motherboard further includes an electrical connection a video graphics conversion module for controlling the chipset and the video graphics conversion interface, when the video graphics conversion module of the baseboard management controller receives a video graphics conversion signal from the display via the video graphics conversion interface as the message output request The video graphics conversion module of the baseboard management controller converts the at least one system error message into a display signal, and outputs the video signal to the display via the video graphics conversion interface. The server of claim 13, wherein the motherboard further comprises: a basic input/output unit, storing the error comparison table, electrically connecting the control chip set, and initializing the video graphics conversion mode of the substrate management controller Grouping, and after the initialization is completed, notifying the baseboard management controller to cause the baseboard management controller to stop outputting the display signal, The basic input/output unit receives the at least one system error code from the control chipset, and accordingly finds the corresponding at least one system error message from the error comparison table, the basic input/output unit causes the at least one system error The message is converted into at least one display signal, and the video graphics conversion module of the baseboard management controller and the video graphics conversion interface of the motherboard are outputted to the display.