JP2010198410A - Server failure prediction system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To prevent a failure of a server in advance by preliminarily predicting the possibility of abnormal conditions in the server. <P>SOLUTION: The server records a log, including a time when an event occurs and the contents of the event, into a log area in time series. A maintenance server monitors a log area and collects a first log, including a time when the power supply of the server is turned on, and a second log, including a time when the server becomes usable. The maintenance server calculates a time being a difference between the first log and the second log, as a time after the power supply of the server is turned on before the server becomes usable. The maintenance server notifies, when the time is equal to or more than a setting time, a maintenance person of the possibility of the abnormal conditions in the server. In a server failure prediction system, in consideration of a state of deterioration due to the fact that the server is used for a long time, and the possibility of the abnormal conditions in the server is predicted in advance. Accordingly, the failure of the server can be prevented in advance by notifying the maintenance person of the possibility of the abnormal conditions in the server. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a server failure prediction system that monitors a server and predicts a failure.

  FIG. 1 shows a general system configuration. The server failure prediction system includes a server 110 and a maintenance server 120. The maintenance server 120 is connected to the server 110. The server 110 includes a storage device 112. The storage device 112 is provided with a log area 113.

  The server 110 records logs 113-1, 113-2,... Including the time when the event occurred and the contents of the event in the log area 113 of the storage device 112 in time series. The maintenance server 120 monitors the log area 113 and assumes that there is a log indicating an error in the event content. Examples of the error include device voltage abnormality and temperature. In this case, the maintenance staff is notified of the abnormality of the server 110.

  Conventionally, it is determined whether or not the server 110 is abnormal for a single event. That is, the maintenance staff is notified only when there is an abnormality. For this reason, when an error occurs, maintenance personnel perform maintenance work for the first time. In maintenance work, devices are replaced or repaired. It is desirable to prevent a server failure by predicting in advance that there may be an abnormality in the server.

  Introduces literature on failure prediction and device monitoring.

  Japanese Patent Application Laid-Open No. 2001-31375 describes a failure prediction system for an external storage device. A failure prediction system for an external storage device includes an external storage device, a customer computer using the external storage device, and a service provider computer connected to the customer computer via a communication network. . The customer computer acquires inspection data relating to the usage status of the external storage device, and transmits the inspection data to the service provider computer via the communication network. The service provider computer predicts a failure of the external storage device based on the inspection data, and transmits the result to the customer computer via the communication network.

  Japanese Patent Laid-Open No. 2004-213621 describes a remote monitoring system. The remote monitoring system receives event information including normality / abnormality of the monitored system, and reports the e-mail including the received event information periodically or irregularly via a network path, and the first means And a second means for monitoring the monitored system and the state of the network path together by receiving an electronic mail.

  Japanese Patent Laid-Open No. 2002-259130 describes an information processing system. The information processing system measures the elapsed time since the start of the operating system, the means for detecting the completion of the start of the operating system, and the start signal for starting the operating system, and the start signal is generated And a means for controlling switching of a boot device to be activated by the operating system based on whether or not the activation completion of the operating system is detected within a predetermined elapsed time.

  Japanese Patent Application Laid-Open No. 2006-236524 describes an image processing apparatus. The image processing apparatus includes a magnetic storage unit that can store a program and image data of the image processing apparatus and has a drive mechanism therein, and a control unit that controls the image processing apparatus. The control means is characterized by transmitting a command to the magnetic storage means in order to acquire failure diagnosis information of the magnetic storage means, and performing failure prediction of the internal drive mechanism of the magnetic storage means based on the acquired failure diagnosis information. .

JP 2001-31375 A JP 2004-213621 A JP 2002-259130 A JP 2006-236524 A

  An object of the present invention is to provide a server failure prediction system capable of preventing a server failure by predicting in advance that there may be an abnormality in the server.

  The server failure prediction system of the present invention includes a server and a maintenance server connected to the server. The server includes a storage device, and records a log including the time when the event occurs and the contents of the event in a time series in the log area of the storage device. The maintenance server includes a monitoring unit, a calculation unit, and a notification unit. The monitoring unit monitors the log area and collects a first log including a time when the server is turned on and a second log including a time when the server becomes available. The calculation unit calculates a first time, which is a difference between the time of the first log and the second log, as the time from when the server is turned on until the server becomes available. When the first time is equal to or longer than the first set time, the notification unit notifies the maintenance staff that there is a possibility that the server has an abnormality.

  In the server failure prediction system of the present invention, the maintenance server has an abnormality in the server by determining whether the time from when the server is turned on until the server becomes available is longer than a set time. Predict whether or not there is a possibility. In other words, taking into account the deterioration caused by the server being used for a long time (degradation of the total startup time of the server due to deterioration of the seek time of the hard disk or deterioration of the response time of the device due to heat, etc.) Predicting that there is a possibility. Therefore, it is possible to prevent a server failure by notifying maintenance personnel that there is a possibility that the server has an abnormality.

FIG. 1 shows a general system configuration. FIG. 2 shows the configuration of the server failure prediction system according to the embodiment of the present invention. FIG. 3 is a flowchart showing the operation of the server failure prediction system according to the embodiment of the present invention. FIG. 4 is a diagram for explaining the operation of the server failure prediction system according to the embodiment of the present invention.

  Hereinafter, a server failure prediction system according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 2 shows the configuration of the server failure prediction system according to the embodiment of the present invention. The server failure prediction system according to the embodiment of the present invention includes a server 10 and a maintenance server 20. The maintenance server 20 is connected to the server 10.

  The server 10 is a computer, and includes a CPU (Central Processing Unit) 11, a storage device 12, and a plurality of devices. Examples of the plurality of devices include a memory such as a hard disk and a chip set.

  The storage device 12 is provided with an area for storing a computer program to be executed by the server 10 and a log area 13. The CPU 11 reads and executes a computer program from the storage device 12 at the time of startup or the like.

  The maintenance server 20 is a computer, and includes a CPU 21, a storage device 22, a display device 27, and a speaker 28.

  The storage device 22 is provided with an area in which a computer program 23 to be executed by the CPU 21 is stored. The CPU 21 reads and executes the computer program 23 from the storage device 22 at the time of startup or the like.

  The computer program 23 includes a monitoring unit 24, a calculation unit 25, and a notification unit 26.

  FIG. 3 is a flowchart showing the operation of the server failure prediction system according to the embodiment of the present invention.

  First, the operation of the server 10 will be described.

  The CPU 11 starts up the server 10 when the user turns on the power of the server 10 (step S1).

  The CPU 11 records the logs 13-1, 13-2,... Including the time when the event occurs and the contents of the event in the log area 13 of the storage device 12 in time series (step S2).

  When the user does not give an instruction to end the startup of the server 10 (step S3-NO), the CPU 11 executes step S2, and when the user gives an instruction to end the startup of the server 10, the server 10 Is finished (step S3-YES).

  Next, the operation of the maintenance server 20 will be described.

  Here, in addition to the operation of the maintenance server 120 described above (hereinafter referred to as error processing), the maintenance server 20 executes the following operation (hereinafter referred to as prediction processing).

  In the error processing, the monitoring unit 24 monitors the log area 13 and it is assumed that there is a log indicating an error in the event content. Examples of the error include device voltage abnormality and temperature. In this case, the notification unit 26 notifies the maintenance staff of the abnormality of the server 10.

  The prediction process will be described.

  The monitoring unit 24 monitors the log area 13 and, as shown in FIG. 2, the first log including the time when the power of the server 10 is turned on by the user (hereinafter referred to as the power-on log 13-1). And a second log including the time when the server 10 becomes available (hereinafter referred to as a start completion log 13-j) (step S11).

  The calculation unit 25 calculates the time between the power-on log 13-1 and the start completion log 13-j as the time from when the power of the server 10 is turned on until the server 10 becomes available (time required for the specific event). A first time Δt (hereinafter referred to as time Δt1), which is a difference, is calculated (step S12).

  The notification unit 26 compares the time Δt1 with a predetermined first set time t (hereinafter referred to as set time t1) (step S13).

  Therefore, when the time Δt1 is less than the set time t1 (step S13—NO), the maintenance server 10 executes step S11.

  On the other hand, when the time Δt1 is equal to or longer than the set time t1 (step S13—YES), the notification unit 26 displays on the display device 27 by text that there is a possibility that the server 10 is abnormal, and this is displayed from the speaker 28. The sound is output and notified to maintenance personnel (step S14).

  When the maintenance server 10 has a specification that ends the prediction process once notified to the maintenance staff, the maintenance server 10 ends the prediction process after executing step S14. Alternatively, the maintenance server 10 executes step S <b> 11 (not shown) when the specification is such that the abnormality of the server 10 is further detected even if the maintenance staff is notified once.

  Thus, in the server failure prediction system according to the embodiment of the present invention, the maintenance server 20 determines whether or not the time Δt1 from when the server 10 is powered on until the server 10 becomes available is equal to or longer than the set time t1. It is predicted in advance whether or not there is a possibility that the server 10 has an abnormality. That is, the server 10 is considered in consideration of the deterioration due to the server 10 being used for a long time (deterioration of the total startup time of the server 10 due to deterioration of the seek time of the hard disk or the response time of the device due to heat). It is predicted in advance that there may be an abnormality. Therefore, it is possible to prevent a failure of the server 10 by notifying maintenance personnel that there is a possibility that the server 10 has an abnormality.

  Here, the maintenance server 20 performs the following operation on a specific device among the plurality of devices.

  The monitoring unit 24 monitors the log area 13 and, as shown in FIG. 4, a third log including the time when the specific device is activated (hereinafter referred to as activation start log 13-x), the specific device A fourth log (hereinafter referred to as a start completion log 13-y) including the time when becomes available is collected (step S11).

  The calculation unit 25 is the difference between the time of the start start log 13-x and the start completion log 13-y as the time from when the specific device starts up until it becomes usable (the time required for the specific event). 2 hours Δt (hereinafter referred to as time Δt2) is calculated (step S12).

  The notification unit 26 of the maintenance server 20 compares the time Δt2 with a predetermined second set time t (hereinafter referred to as set time t2) (step S13).

  Therefore, when the time Δt2 is less than the set time t2 (step S13—NO), the maintenance server 10 executes step S11.

  On the other hand, when the time Δt2 is equal to or longer than the set time t2 (step S13—YES), the notification unit 26 uses a character display to indicate that there is a possibility that the specific device has an abnormality as a possibility that the server 10 has an abnormality. 27, which is output by sound from the speaker 28 and notified to maintenance personnel (step S14).

  When the maintenance server 10 has a specification that ends the prediction process once notified to the maintenance staff, the maintenance server 10 ends the prediction process after executing step S14. Alternatively, the maintenance server 10 executes step S <b> 11 (not shown) when the specification is such that the abnormality of the server 10 is further detected even if the maintenance staff is notified once.

  As described above, in the server failure prediction system according to the embodiment of the present invention, it is determined whether or not the time Δt2 from when the specific device is activated until it becomes available is greater than or equal to the set time t2. Predict whether there is a possibility of abnormality. Therefore, it is possible to prevent a failure of the server 10 by notifying maintenance personnel that there is a possibility that the specific device has an abnormality.

10 servers,
11 CPU,
12 storage devices,
13 Log area,
13-1, 13-2, 13-j, 13-x, 13-y log,
20 maintenance server,
21 CPU,
22 storage devices,
23 computer program,
24 monitoring unit,
25 calculation unit,
26 Notification section,
27 display device,
28 speakers,
110 servers,
112 storage device,
113 log area,
113-1, 113-2 logs,
120 maintenance server,

Claims (13)

A server comprising a storage device, and recording a log including the time when the event occurred and the contents of the event in a time series in the log area of the storage device;
A maintenance server connected to the server,
The maintenance server
A monitoring unit that monitors the log area and collects a first log including a time when the server is powered on and a second log including a time when the server is available;
A calculation unit that calculates a first time that is a difference between the time of the first log and the second log as a time from when the server is turned on until the server becomes available;
A server failure prediction system comprising: a notification unit that notifies a maintenance person that there is a possibility of an abnormality in the server when the first time is equal to or longer than a first set time. The server further comprises a plurality of devices,
The monitoring unit monitors the log area, and includes a third log including a time when a specific device of the plurality of devices is activated, and a time including a time when the specific device is available. 4 logs and collect
The calculation unit calculates a second time, which is a difference between the time of the third log and the fourth log, as the time from when the specific device is activated until it becomes usable,
The notifying unit notifies a maintenance person that there is a possibility that the specific device may be abnormal as a possibility that the server has an abnormality when the second time is equal to or longer than a second set time. Server failure prediction system described in 1. The maintenance server
A display device;
3. The server failure prediction system according to claim 1, wherein the notification unit displays on the display device by characters that there is a possibility of abnormality in the server, and notifies a maintenance staff. The maintenance server
A speaker,
The server failure prediction system according to any one of claims 1 to 3, wherein the notification unit outputs a sound from the speaker that there is a possibility that the server is abnormal, and notifies a maintenance staff. A maintenance server connected to a server that records a time series of logs including the time when an event occurs and the contents of the event in its own log area,
A monitoring unit that monitors the log area and collects a first log including a time when the server is powered on and a second log including a time when the server is available;
A calculation unit that calculates a first time that is a difference between the time of the first log and the second log as a time from when the server is turned on until the server becomes available;
A maintenance server comprising: a notification unit for notifying maintenance personnel that there is a possibility of an abnormality in the server when the first time is equal to or longer than a first set time. The monitoring unit monitors the log area, and includes a third log including a time when a specific device of a plurality of devices of the server is activated, and a time when the specific device becomes available. 4th log including
The calculation unit calculates a second time, which is a difference between the time of the third log and the fourth log, as the time from when the specific device is activated until it becomes usable,
The notifying unit notifies maintenance personnel that there is a possibility that the specific device is abnormal as a possibility that the server has an abnormality when the second time is equal to or longer than a second set time. The maintenance server described in. A display device;
The maintenance server according to claim 5 or 6, wherein the notification unit displays on the display device by text that there is a possibility that the server has an abnormality, and notifies a maintenance staff. A speaker,
The maintenance server according to any one of claims 5 to 7, wherein the notification unit outputs a sound from the speaker that there is a possibility of an abnormality in the server, and notifies a maintenance staff. A method of using a computer connected to a server that records a log including a time when an event occurs and contents of the event in a time series in its own log area,
Monitoring the log area and collecting a first log including a time when the server is powered on and a second log including a time when the server is available;
Calculating a first time, which is a difference between the time of the first log and the second log, as a time from when the server is turned on until the server becomes available;
A server failure prediction method comprising: notifying maintenance personnel that the server may be abnormal when the first time is equal to or longer than a first set time. A third log including a time when a specific device of a plurality of devices of the server is activated by monitoring the log area; and a fourth log including a time when the specific device becomes available; Collecting steps,
Calculating a second time, which is a difference between the time of the third log and the fourth log, as the time from when the specific device is activated until it becomes available;
And a step of notifying a maintenance staff that there is a possibility that the specific device has an abnormality as a possibility that the server has an abnormality when the second time is equal to or longer than a second set time. 9. The server failure prediction method according to 9. The step of notifying the maintenance staff includes:
The server failure prediction method according to claim 9 or 10, wherein a message indicating that there is a possibility of an abnormality in the server is displayed on the display device and notified to maintenance personnel. The step of notifying the maintenance staff includes:
The server failure prediction method according to any one of claims 9 to 11, wherein the server is output by sound from the speaker that there is a possibility of abnormality in the server, and is notified to maintenance personnel.   The computer program which makes the said computer perform the server failure prediction method in any one of Claims 9-12.

Images