CN1581079B - Automatic restarting method and system for down of network server - Google Patents

Abstract

A method and a system for automatically restarting the down of a network server can be used for the network server, and when the down of the network server occurs, the network server automatically executes a reactivation program to restore the normal operation state. The automatic network server downtime restarting method and system features that one SMI processing program is triggered by the monitoring timer in the I/O control chip in the network server to trigger the main computer unit of the network server to execute one re-activating program automatically. The characteristic can lead the network server to have the downtime automatic restart function without spending extra cost to purchase software and hardware facilities, and the downtime automatic restart function can not be executed because of the downtime of the operating system.

Description

Method and system for automatically restarting network server when downtime

technical field

The present invention relates to a computer network server technology, in particular to a method and system for automatically restarting a network server when the network server is down. The program (reboot) restores the normal operating state of the network server that has been down.

Background technique

The network server provides users with specific service functions through the network system. If the network server goes down during operation, all users will be unable to use the service functions provided by it. Therefore, whenever a downtime occurs in the network server, it must be rebooted immediately, so that the downtime network server can be restored to a normal operating state.

A commonly used method is that when a network server is found to be down, the network system administrator manually executes a restart procedure for the network server where the downtime occurs. The disadvantage of this method is that after the network server goes down, it may be idle for a period of time before being detected by the network system administrator and performing a restart procedure, so users cannot use it during this idle time. The service function provided by the web server in the downtime state.

A solution to the above-mentioned problem is to set an automatic restart system in the network server, which can automatically answer whether the network server is down; and when the network server is down, the host unit of the network server can automatically execute A reactivation program, whereby the operating system is reloaded to restore normal operation to a network server that has been down.

FIG. 1 shows the basic system architecture of an existing system 100 for automatically restarting a network server when it fails. As shown in the figure, this existing network server shutdown automatic restart system 100 is carried to the network server 10, and the host unit 20 of the network server 10 (that is, the general term of the central processing unit and all related hardware facilities) executes a server Dedicated operating system 30; and the basic architecture of this existing network server downtime automatic restart system 100 includes: (a) a downtime status response module 110; (b) a monitoring timer (watchdog timer) 120; and ( c) an IRQ processing module 130 . In practical implementation, the existing system 100 for automatically restarting when the network server is down is made into a circuit card, and the circuit card is connected to the network server 10 to provide the function of automatically restarting when the network server is down.

Downtime status response module 110 regularly reflects whether the operating system 30 of the network server 10 is in a normal operating state, that is, every predetermined time period, responding to the operation information sent by the operating system 30 in the network server 10; The received operation information is transmitted to the monitoring timer 120 synchronously.

The monitoring timer 120 can start to execute a timing program of a predetermined length of time when receiving the operation information sent by the above-mentioned downtime status response module 110 each time, for example, counting from zero to 59 seconds; Before the cycle, when receiving the next operation information sent by the above-mentioned downtime status response module 110 again, it will be reset to zero and count again; After a predetermined timing period, for example, 60 seconds (that is, the watchdog timer 120 overflows), a system abnormality signal is sent to the IRQ processing module 130 .

The IRQ processing module 130 responds to the system abnormality signal sent by the above-mentioned monitoring timer 120, and sends an IRQ restart signal RESET to the host unit 20 of the network server 10, so that the host unit 20 automatically executes a restart program, that is, reloads the operating system 30 , allowing the network server 10 in the downtime state to restore the normal operation state.

A disadvantage of the above-mentioned existing network server automatic restart system 100 is that it is an external circuit card, so users need to spend extra cost to purchase this circuit card, which increases the operating cost.

In addition, because the above-mentioned existing automatic restart system 100 of the network server when it is down is based on the IRQ signal to complete the automatic restart function when it is down, if the operating system therein also has a downtime and cannot process the IRQ signal, then it cannot achieve downtime. Automatic restart function after shutdown.

Contents of the invention

In order to overcome the shortcomings of the above-mentioned prior art, the main purpose of the present invention is to provide a new method and system for automatically restarting a network server when it is down. It has the function of automatic restart after downtime.

Another object of the present invention is to provide a new method and system for automatically restarting a network server when it is down, which does not need to use the IRQ signal to perform the function of automatically restarting when it is down.

A kind of network server downtime automatic restarting method of the present invention can be used in network server, when this network server downtime situation occurs, make this network server automatically execute a reactivation procedure; This network server downtime automatic restarting method at least includes: Reflect whether the operating system of the network server is in a normal operating state at intervals of a predetermined period of time, and transmit this operation information synchronously; each time an operation information is received, it immediately starts to execute a timing program of a predetermined period of time, and again When an operation message is received, it is reset to zero and timed again; if no operation message is received, when the timing program reaches the predetermined timing length, an SMI (System Management Interrupt, System Management Interrupt) signal is sent; and the SMI signal is responded A restart signal is sent to make the network server automatically execute a restart procedure.

A kind of network server downtime automatic restarting system of the present invention can be used for network server, when this network server downtime situation occurs, make this network server automatically execute a reactivation program; Network server downtime automatic restarting system at least includes: a downtime A machine status response module, which can reflect whether the operating system of the network server is in a normal operating state; a monitoring timer, which can start to execute a predetermined time length every time the operation information sent by the downtime status response module is received Timing program, and when the operation information sent by the downtime status response module is received again, it is reset to zero and timed again; if the operation information is not received, when the timing program reaches the predetermined timing length, a system abnormality alarm is sent signal; and an SMI processing module, which can respond to the system abnormal signal sent by the monitoring timer and send a system management interrupt (System Management Interrupt; SMI) restart signal to execute the SMI judgment processing program, so as to judge whether the restart signal is sent by the SMI processing module , if it is issued by the module, the network server automatically executes a restart procedure through the BIOS restart procedure.

The feature of network server downtime automatic restart method and system of the present invention is that directly adopts the monitoring timer in the I/O control chip (such as Southbridge chip or Super I/O chip) in the network server to trigger the SMI processing program (and It is not the IRQ processing program adopted by the prior art), and the host unit of the network server is triggered by this SMI processing program to make it automatically execute a reactivation program.

Compared with the prior art, the present invention directly utilizes the existing hardware facilities (that is, south bridge chip or Super I/O chip) in the network server, and only modifies the BIOS in the network server (that is, writes the SMI judgment processing program ), so that users do not need to spend additional costs to purchase software and hardware facilities, so that the network server can have the automatic restart function of the downtime. In addition, because the present invention does not need to use the IRQ signal to perform the automatic restart function of the downtime, it will not Due to the downtime of the operating system, the automatic restart function cannot be executed.

Description of drawings

FIG. 1 is a schematic diagram of a system architecture, showing the basic architecture of an existing network server automatic restart system when down; and

FIG. 2 is a schematic diagram of the system architecture, showing the basic architecture of the system for automatically restarting the network server in case of failure of the present invention.

Detailed ways

Example

FIG. 2 shows the basic structure of the system for automatically restarting the network server when the network server crashes (such as the part included in the dotted box indicated by the reference number 200 ) of the present invention. As shown in the figure, in actual application, the automatic restart system 200 of the network server of the present invention is used for the network server 10, and the host unit 20 of the network server 10 (that is, the general name of the central processing unit and all related hardware facilities thereof) ) executes the operating system 30 dedicated to the server. The automatic restart system 200 of the network server of the present invention can automatically cause the network server 10 to execute a reactivation program (reboot) when the host unit 20 or the operating system 30 of the network server 10 is in a downtime condition, so that the downtime occurs The web server 10 resumes normal operation.

The basic framework of the object-oriented component model (object-oriented component model) of the automatic restart system 200 of the network server downtime of the present invention at least includes: (a) a downtime status response module 210; (b) a monitoring timer 220; and (c) An SMI processing module 230 .

The automatic restarting system of the network server of the present invention includes: a downtime status response module 210, which is used to regularly respond to whether the operating system 30 of the network server 10 is in a normal operating state, that is, to respond to the network server 10 at intervals of a predetermined length of time. The information sent by the operating system 30 in the middle;

The implementation mode of watchdog timer 220 is to directly utilize the I/O control chip in the network server 10, such as Southbridge (Southbridge) or Super I/O chip, wherein built-in watchdog timer (watchdog timer); And this watchdog timer 220 can immediately start to execute a timing program of a predetermined time period every time an operation information transmitted by the above-mentioned downtime response module 210 is received, such as counting from zero to 59 seconds; if the predetermined timing period is reached Before, when receiving the next normal operation information sent by the above-mentioned downtime status response module 210 again, it will be reset to zero and count again; if the next operation information cannot be received, the timing program of the monitoring timer 220 will be After the predetermined timing period is exceeded, for example, at the 60th second (that is, the watchdog timer 220 overflows), a system abnormality signal is sent to the SMI processing module 230 .

The embodiment of SMI processing module 230 is to utilize the input and output control chip in the network server 10, such as South Bridge (Southbridge) or Super I/O chip, wherein the built-in SMI signal module, and via the BIOS (Basic A judging module in Input/Output System) judges whether its SMI signal is sent by the monitoring timer 220, if the SMI signal sent by the monitoring timer 220, then the BIOS sends a restart signal RESET to the host unit 20, so that the host unit 20 automatically executes a restart procedure, that is, the operating system 30 is reloaded, so that the network server 10 in the downtime situation can be restored to the normal operation state again.

The network server network server downtime automatic restart method of the present invention at least includes: downtime status response module 210 continues every predetermined time period, just reflects whether the operating system 30 of the network server 10 is in the operating state; and the received information Synchronously sent to the monitoring timer 220; the monitoring timer 220 can immediately start to execute a predetermined time period timing program when receiving an operation information sent by the downtime status response module 210 each time; Before the timing period, when receiving the next operation information sent by the downtime status response module 210 again, it will be reset to zero and count again; if the next operation information sent by the downtime status response module 210 is not received, the monitoring timing The register will overflow, and then send an exception signal to the SMI processing module 230 . The SMI processing module 230 can respond to the abnormal signal sent by the watchdog timer 220, and judge whether it is sent by the watchdog timer 220 by the BIOS, and if so, send a restart signal RESET to the host unit 20, so that the host unit 20 automatically executes a reset procedure, That is, the operating system 30 is reloaded, so that the network server 10 that has crashed can be restored to normal operation.

In a word, the present invention provides a novel method and system for automatically restarting a network server when it is down, which can be applied to a network server, and is used to automatically execute the - Reactivate the program. The feature of network server downtime automatic restart method and system of the present invention is to directly utilize the monitor timer in the I/O control chip (such as Southbridge chip or Super I/O chip) in the network server to trigger an SMI process program (instead of the IRQ processing program adopted in the prior art), the host unit of the network server is triggered by the SMI processing program to automatically execute a reactivation program. Compared with the prior art, because the present invention directly utilizes the existing hardware facilities (that is, Southbridge chip or Super I/O chip) in the network server, and only modifies the BIOS in the network server (that is, writes SMI Judgment processing program), so users do not need to spend extra costs to purchase hardware and software facilities, so that the network server can automatically restart when it is down. In addition, since the present invention does not need to use the IRQ signal to perform the auto-restart function when the system is down, the auto-restart function will not be disabled due to the downtime of the operating system. Therefore, the present invention has better progress and practicability than the prior art.

Claims (10)

1. A method for automatically restarting a network server when it is down, can be used for a network server, and when the network server is down, the network server is automatically executed a reactivation program; it is characterized in that the method for automatically restarting the network server when it is down Include at least: Reflect whether the operating system of the network server is in a normal operating state at intervals of a predetermined period of time, and transmit this operating information synchronously; Every time an operation message is received, it starts to execute a timing program with a predetermined length of time, and is reset to zero when receiving an operation message again; if the timing program reaches the predetermined timing length and does not receive it again When the operation information is received, an SMI signal is issued; and, A restart signal is sent in response to the SMI signal, so that the network server automatically executes a restart procedure. 2. The method for automatically restarting a network server when it fails according to claim 1, wherein the timing program is executed by a monitoring timer built in the I/O control chip. 3. The network server downtime automatic restart method as claimed in claim 2, wherein the I/O control chip is a Super I/O chip. 4. The method for automatically restarting a network server when it is down as claimed in claim 2, wherein the I/O control chip is a south bridge chip. 5. A network server shutdown automatic restart system can be used for a network server. When the network server has a shutdown situation, the network server is automatically executed to reactivate the program; it is characterized in that the network server shutdown automatic restart system is at least include: A downtime status response module, which can reflect whether the operating system of the network server is in a normal operating state; A monitoring timer, each time it receives the operation information sent by the downtime status response module, it starts to execute a timing program with a predetermined length of time, and is reset when the operation information sent by the downtime status response module is received again reset to zero; When the timing program reaches a predetermined timing length without receiving any operation information again, it sends out a system abnormality signal; and An SMI processing module can send a system management interrupt restart signal in response to the system abnormal signal sent by the monitoring timer to execute the SMI judgment processing program, so as to judge whether the restart signal is sent by the SMI processing module, and if it is sent by the module, then pass The BIOS reboot program enables the network server to automatically execute a reboot program. 6. The network server automatic restart system according to claim 5, wherein the monitoring timer is a function module built in the I/O control chip. 7. The automatic restart system of the network server according to claim 6, wherein the I/O control chip is a SuperI/O chip. 8. The automatic restart system of the network server according to claim 6, wherein the I/O control chip is a south bridge chip. 9. The system for restarting automatically when the network server is down as claimed in claim 5, wherein the SMI processing module is a functional module in the I/O control chip of the network server. 10. The system for restarting automatically when the network server fails according to claim 5, wherein the SMI processing module is a functional module in the basic input/output system of the network server.