JP2012178014A - Failure prediction/countermeasure method and client server system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress any damage due to unpredictable failure occurrence by calculating the degree of deterioration by collecting the information of a device, and executing failure countermeasures in an early stage according to the degree of deterioration.SOLUTION: This client server system includes a management server 100 connected via a network 500 to a maintenance object device 200, and the management server 100 includes management means 70 for acquiring the information of the maintenance object device in a fixed cycle from the maintenance object device, and for managing the device. The management means 70 is configured to calculate the progress of deterioration of the maintenance object device on the basis of the acquired information of the maintenance object device, and to, when the calculated progress of deterioration is turned into a predetermined progress, or the information acquired from the maintenance object device shows a predetermined event, execute failure countermeasures to the maintenance object device. The failure countermeasures are executed for mirroring with an auxiliary device 210 and operation switching to the auxiliary device 210.

Description

  The present invention relates to a failure prediction / countermeasure method and a client / server system, and in particular, it is possible to predict a failure of a device to be maintained such as an information processing device in advance and to implement a countermeasure against the failure. The present invention relates to a countermeasure method and a client server system to which the method is applied.

  As a conventional technique for predicting a failure of an apparatus, for example, a technique described in Patent Document 1 is known. This prior art predicts a failure of the maintenance target device based on information on the operation status of the maintenance target device periodically transmitted from a monitoring terminal connected to the maintenance target device, and confirms that the failure time is near. It detects and notifies the user.

JP 2009-217770 A

  The conventional technology for predicting the failure of the maintenance target device described above only informs the user that the failure time is near, so there is a risk that a failure will occur before the user actually takes action on the device. There is a problem that there is. In addition, the above-described conventional technology does not take into account a countermeasure when a failure occurs earlier than predicted, and in such a case, there is a problem that there is a risk of data loss or system down. ing.

  The object of the present invention is to solve the above-mentioned problems of the prior art, calculate the degree of deterioration by collecting information on the apparatus, take measures against the failure early according to the degree of deterioration, and by the occurrence of a failure that makes predictions faster It is an object of the present invention to provide a failure prediction / countermeasure method and a client server system that can suppress damage.

  According to the present invention, the object is to predict a failure of a device to be maintained in advance and implement a countermeasure against the failure. In the device failure prediction / measure method, the device is connected to the device to be maintained through a network. The management server has management means for acquiring information on the apparatus from the apparatus to be maintained at a constant cycle and managing the apparatus, and the management means Based on the information of the device to be maintained, the deterioration progress of the device to be maintained is calculated, and when the calculated deterioration progress becomes a predetermined progress, or the device to be maintained When the information acquired from the above indicates a predetermined event, this is achieved by implementing a countermeasure against the device to be maintained.

  According to the present invention, when it is determined that there is a high risk of failure of a device to be maintained such as an information processing device, it is possible to take measures to avoid the failure. In addition to the reduction, it is possible to avoid a situation in which the apparatus is broken even though the risk of failure is predicted.

It is a block diagram which shows the structure of the client server system by one Embodiment of this invention. It is a block diagram which shows the function structure of the blade module and management server in the client server system by embodiment of this invention. It is a figure explaining the path | route which the management server acquires the information used in order to determine deterioration of a blade module and to predict a failure. 6 is a sequence chart (part 1) illustrating an example of processing for determining deterioration of a monitoring target blade module and taking a countermeasure against a failure in the client server system according to the embodiment of the present invention. It is a sequence chart (the 2) explaining the example of the process which determines deterioration of a monitoring object blade module, and performs a countermeasure against a failure in the client server system by embodiment of this invention. It is a figure explaining the information used for the primary deterioration determination for a primary failure countermeasure. It is a figure explaining the information of the event which implements a secondary failure countermeasure. It is a sequence chart (the 1) explaining other examples of processing which judges degradation of a monitoring object blade module and performs a countermeasure against failure in a client server system by an embodiment of the present invention. It is a sequence chart (the 2) explaining the other example of the process which determines deterioration of a monitoring object blade module, and performs a countermeasure against failure in the client server system by embodiment of this invention.

  Embodiments of a failure prediction / measure method and a client server system according to the present invention will be described below in detail with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration of a client server system according to an embodiment of the present invention.

  The client server system according to the embodiment of the present invention shown in FIG. 1 includes a server 300 including a plurality of blade modules 200, a management server 100 to which a management terminal 600 used by an administrator is connected, and a plurality of clients used by a user. A terminal 400 is connected to a network 500 such as an intranet. Each of the plurality of blade modules 200 is an information processing apparatus having at least an HDD 30, a RAM 40, and a CPU, and is connected to the client terminal 400 via the network 500 and has a function of providing various information processing services to the user. Have. Similarly to the blade module 200, the management server 100 is an information processing apparatus that includes an HDD, a RAM, a CPU, and the like, acquires information from the blade module 200, and manages and controls the blade module 200. Has the function to perform. The client terminal 400 is an information processing apparatus as a thin client that includes at least an input device such as a keyboard, a display device, a memory, and a CPU. The client terminal 400 is connected to the blade module 200 and is connected to the blade module 200 in various ways. The information processing is executed and the execution result is received.

  FIG. 2 is a block diagram showing functional configurations of the blade module 200 and the management server 100 in the client server system according to the embodiment of the present invention. The blade module in the client server system according to the embodiment of the present invention includes a blade module to be monitored and a spare blade module. FIG. 2 shows one blade module 200 to be monitored and one spare blade module. 210 is shown.

  The server 300 in the client server system according to the embodiment of the present invention includes a plurality of monitoring target blade modules 200 and includes a number of spare blade modules 210 smaller than the number of monitoring target blade modules 200. The monitoring target blade module 200 is an information processing apparatus that is used as a personal computer by being accessed from the outside, and is an apparatus that is normally used when operating a server using a plurality of blade modules. The spare blade module 210 is an information processing apparatus that can be used as a personal computer by being accessed from the outside, and is operated by applying the failure prediction / countermeasure method according to the present invention to a system using a plurality of blade modules. This is a device used when taking measures against failure of the monitoring target blade module 200.

  The aforementioned monitoring target blade module 200 and spare blade module 210 have the same functional configuration as shown in FIG. That is, each of these modules includes Agent bodies 10 and 11, OS 20 and 21, HDD (or SSD) 30 and 31, module main bodies 45 and 46 having ECC-attached RAMs 40 and 41, monitoring control devices BMCs 50 and 51, Communication devices NIC60 and 61 are provided. The management server 100 includes a control unit 75 having a Manager 70 and a communication device NIC60. The blade modules 200 and 210 and the management server 70 operate by program control, for example, and are connected to each other via a network 500 such as an intranet.

  The management server 100 is an information processing apparatus for managing the blade modules 200 and 210. By using the installed Manager 70, the management server 100 acquires, manages, and manages the information of the managed blade modules 200 and 210. The blade modules 200 and 210 can be controlled. The Manager 70 is software for managing the blade modules 200 and 210. The Manager 70 is configured to communicate with the BMCs 50 and 51 of the blade modules 200 and 210. It has a function for requesting various types of information and a function for recording and managing information of the blade modules 200 and 210.

  The Agents 10 and 11 of the blade modules 200 and 210 are software that assists in sending and receiving information between the Manager 70 of the management server 100 and the blade modules 200 and 210, and are installed and used in the blade modules 200 and 210. The The Agents 10 and 11 have a function of notifying the manager 70 of the power state transition of the blade modules 200 and 210, and a function of acquiring information of the blade modules 200 and 210 in response to a request from the Manager 70 and transmitting the information to the Manager 70. ing.

  HDDs (SSD) 30 and 31 are auxiliary storage devices built in the blade modules 200 and 210, and have S.M.A.R.T. information which is self-diagnosis information. The RAMs 40 and 41 with ECC are main storage devices built in the blade modules 200 and 210, and can detect and correct 1-bit error out of 64 bits, and 2 bits out of 64 bits. It has a function that can detect errors. When the error is detected, the RAM with ECC 40 and 41 transmits the information to the OS 20 and 21 as an ECC error log.

  The NICs 60, 61, and 62 are devices that control communication between information processing apparatuses via the network 500. The NICs 60 and 61 in the blade modules 200 and 210 having the BMCs 50 and 51 are connected to the blade module main bodies 45 and 46, respectively. Control of communication between each of the BMCs 50 and 51 and the management server 400 and the client terminal 400 as an external information processing apparatus is performed.

  The BMCs 50 and 51 are monitoring and control devices incorporated in the blade modules 200 and 210. Only the BMCs 50 and 51 operate in the blade modules 200 and 210 with independent power supplies, and the power control function of the blade modules 200 and 210, It has a function of acquiring voltage values and temperature values of the substrates of the blade modules 200 and 210. The BMCs 50 and 51 also have a function of performing power control and information acquisition for the own blade modules 200 and 210 in accordance with instructions from the external information processing apparatus via the network 500 by using IPMI commands. The OSs 20 and 21 of the blade modules 200 and 210 have identifiers of information processing apparatuses in the network 500 called IP addresses, but the BMCs 50 and 51 are unique from the OSs 20 and 21 of the blade modules 200 and 210. IP address.

  FIG. 3 is a diagram for explaining a path through which the management server 100 acquires information used for determining deterioration of the blade module 200 and predicting a failure.

  When the management server 100 predicts a failure of the monitoring target blade module 200, the management server 100 (1) of the monitoring target blade module 200 (1) S.D. M.M. A. R. T.A. Information, (2) ECC error log of RAM 40 with ECC, (3) hardware monitor log, (4) cumulative execution number of forced power-off / forced reset of blade module 200 performed by Manager 70 of own management server 100, ( 5) Use the voltage and temperature of the board, and (6) the cumulative startup time.

  Then, the management server 100 (1) S. of HDD (SSD) 30. M.M. A. R. T.A. Information, (2) ECC error log of RAM 40 with ECC, (3) Information of hardware monitor log is acquired from Manager 70 to Agent 10 and acquired via Agent 10, (5) Board Each information of voltage / temperature and (6) cumulative activation time is acquired from the BMC 50 by communicating from the Manager 70 to the BMC 50. Also, (4) the cumulative number of executions of forced power OFF / forced reset of the blade module 200 performed by the Manager 70 of the own management server 100 is performed by the Manager 70 itself, and is recorded by the Manager 70 itself.

  4 and 5 are sequence charts for explaining an example of processing for determining the deterioration of the monitored blade module 200 and taking a countermeasure against the failure in the client server system according to the embodiment of the present invention. FIG. FIG. 7 is a diagram for explaining information used for primary deterioration determination, and FIG. 7 is a diagram for explaining information on events for implementing secondary failure countermeasures. These will be described next. The processing of the sequence shown in FIGS. 4 and 5 to be described here is an example of processing for taking measures against failure in two stages from the normal operation, and is a series of processing, and therefore, FIG. 4 and FIG. 5 are continuous. It will be explained as being.

(1) Now, it is assumed that the user connects the client terminal 400 to one of the blade modules of the server 300 and uses the blade module. In this case, the blade module in use becomes the monitoring target blade module 200, and the power of the spare blade module 210 is turned off (sequence A1, A2).

(2) In the above-described state, the Manager 70 of the management server 100 requests the monitored blade module 200 to transmit each piece of information described above with reference to FIG. Information is acquired, and primary degradation determination of the monitoring target blade module 200 is performed based on the information (sequence A3 to A6).

  In the process of performing the primary deterioration determination of the sequence A6, the manager 70 of the management server 100 uses the information shown in FIG. 6, so the information shown in FIG. 6 will be described here.

  The Manager 70 of the management server 100 acquires the information in the column indicated as information on the blade module in FIG. 6 and the value of the information in the column indicated as the value from the monitoring target blade module 200. Then, the Manager 70 compares the threshold value of the information in the column indicated as the threshold value held by the Manager 70 with the value of the acquired information, and the result of the comparison indicates that the information used for the deterioration degree determination It is determined whether the variable in the column indicated as “0” is set to “0” or “1”.

As an example of information of the blade module acquired by the Manager 70 of the management server 100 from the monitored blade module 200 and information held by itself, as shown in FIG.
4. Included in SMART information Substituted number of bad sectors, value Vs5, threshold value Ts5,
Included in SMART information Magnetic head seek error rate, value Vs7, threshold Ts7,
Included in SMART information Power ON / OFF count, value Vs12, threshold Ts12,
193 included in SMART information. Load / unload count, value Vs193, threshold Ts193,
196. included in SMART information. Sector replacement processing occurrence count, value Vs196, threshold value Ts196,
197 included in SMART information. Number of sectors waiting for substitution processing, value Vs197, threshold value Ts197,
198. Included in SMART information. Number of unrecoverable sectors, value Vs198, threshold Ts198,
Cumulative number of forced power off / forced reset, value Vp, threshold Tp,
Number of 1-bit ECC error logs, value Ve1, threshold Te1,
Substrate voltage, value Vv, threshold TvL (minimum voltage threshold), TvH (maximum voltage threshold),
Substrate temperature, value Vt, threshold Tt,
Cumulative startup time, value Va, threshold Ta
There is.

The Manager 70 of the management server 100 compares the value of each information described above with a threshold value, and the F value 0 or 1 according to the conditional expression shown in the condition column where F = 0 and the condition column where F = 1. Is determined as the value of the variable in the column indicating the variable of each information. The variables of each information in the column shown as variables are F, Fs5, Fs7, Fs12, Fs193, Fs196, Fs197, Fs198, Fp, Fe1,
Fv, Ft, and Fa, which are determined by the manager 70 of the management server 100 to the value 0 or 1.

  Then, the Manager 70 of the management server 100 determines the weights W = Ws5, Ws7, Ws12, Ws193, Ws196, Ws197, Ws198, Wp, and the like that are determined in advance for each variable F according to the importance of each information in the determination process of the sequence A6. Multiplying We1, Wv, Wt, Wa (0 <W ≦ 1), and summing the obtained values, the degree of deterioration E is calculated by equation (1), and whether E exceeds the primary deterioration judgment value TE1 It is determined whether or not to deal with the primary failure countermeasure described later.

E = Σ (F ・ W) …… (1)
(3) When the manager 70 of the management server 100 determines in the determination of the sequence A6 that the deterioration progress level E does not exceed the primary deterioration determination value TE1 and the primary failure countermeasure is not performed, the sequence from the sequence A3 Returning to the process, the operation from the process requesting information from the monitored blade module 200 is repeated, and the primary failure occurs when the deterioration progress E exceeds the primary deterioration determination value TE1 and satisfies the primary deterioration determination condition. In order to take countermeasures, the state transits to a standby state (sequence A8).

(4) The Manager 70 of the management server 100 terminates the use of the monitoring target blade module 200 that is being used by the user, and the Agent 10 of the monitoring target blade module 200 from the monitoring target blade module 200 in the above-described primary failure countermeasure standby state. When a power-off notification is received, in order to determine that the received power-off notification is not due to a restart or the like, the system continues to wait for a certain period of time, and if a power-on notification from the monitored blade module 200 is received during the standby No action is taken and the primary failure countermeasure standby state is continued until a power OFF notification is received again. If the manager 70 does not receive a power-on notification during standby, the manager 70 starts mirroring the monitored blade module 200 and the spare blade module 210 as a countermeasure against the primary failure (sequence A7, A9 to A11).

(5) The Manager 70 communicates with the BMCs 50 and 51 of the monitoring target blade module 200 and the spare blade module 210, issues a power ON command by IPMI, and turns on the power of the monitoring target blade module 200 and the spare blade module 210. Operation is performed (sequences A12 to A14).

(6) Thereafter, when Manager 70 receives the blade module power ON notification from the Agents of both blade modules, Manager 70 transmits a data replication command to both blade modules 200 and 210 to start mirroring (sequences A15 to A17). ).

(7) Upon receiving the data replication command for starting mirroring, the monitoring target blade module 200 uses the HDD (SSD) of the spare blade module 210 to store data in an area other than the system area of the HDD (SSD) 30 in the own module 200. Duplicate to 31. At the start of replication, the Agent 10 of the monitoring target blade module 200 notifies the manager 70 of the start of replication (sequence A18, A19).

(8) When the replication of the data in the area other than the system area is completed, the Agent 11 of the spare blade module 210 notifies the Agent 10 of the monitoring target blade module 200 of the completion of the replication (sequence A20).

(9) After transmission / reception of the data replication completion notification, both blade modules 200 and 210 start restarting, notify the power supply OFF to the Manager 70, temporarily turn off the power, and then turn on the power (sequence) A21-A26).

(10) The monitored blade module 200 replicates the data in the system area to the spare blade module 210 before the OS 20 of the own module and the OS 21 of the spare blade module 210 are activated, and thereafter the HDD (SSD) of the monitored blade 200 Mirror setting is performed so that data writing to 30 is performed in the spare blade 210 with exactly the same contents (sequences A27 and A28).

(11) When the above processing is completed, the Agent 11 of the spare blade module 210 notifies the Agent 10 of the monitoring target blade module 200 of the completion of the processing, starts the OS, and completes the mirroring processing (sequence A29). .

(12) Both blade modules 200 and 210 notify the manager 70 that the power is ON, and thereafter, the mirroring of the HDD (SSD) 30 and 31 of both blade modules 200 and 210 always starts to have the same contents. In the mirroring state (sequences A30 to A32, A34).

(13) The processing of the primary failure countermeasure has been completed by the processing described above, and the Manager 70 of the management server 100 continues to each of the monitored blade modules 200 for which the primary failure countermeasure has been completed, at a constant cycle. Information is acquired and secondary deterioration determination is performed based on the acquired information. This secondary deterioration determination process is performed when the value of the deterioration progress E described above becomes equal to or greater than the secondary deterioration determination value TE2 (TE2> TE1), or when the event shown in FIG. 7 occurs. This is a determination process in which the Manager 70 determines to take a secondary failure countermeasure described later (sequences A33, A35 to A37).

  In the process of performing the secondary deterioration determination of the sequence A37, the manager 70 of the management server 100 uses the event information shown in FIG. 7, and therefore the event information shown in FIG. 7 will be described here.

  When the manager 70 of the management server 100 has an event defined by the information in the column indicated as information acquired from the monitored blade module 200 and the content in the column indicating the content of the information, secondary degradation has occurred. It is determined that secondary countermeasures will be taken.

The event is shown in FIG.
1. When the ECC error information is detected as a 2-bit ECC error,
2. Included in SMART information When the reading error rate information exceeds the predetermined threshold (when SMART error log occurs),
3. Included in SMART information When the hard disk processing capacity information exceeds the preset threshold (when a SMART error log occurs),
4). 2. Included in SMART information When the spin-up time information exceeds a predetermined threshold (when a SMART error log occurs),
5. 4. Included in SMART information When the information on the number of defective sectors that have been replaced has increased over the past several times,
6). 196. included in SMART information. When the information on the number of sector replacement processing occurrences increases in value over the past several times of information acquisition,
7). 198. Included in SMART information. This is the time when the information on the number of unrecoverable sectors increases over the past several times of information acquisition.

(14) The manager 70 of the management server 100 does not have the degradation progress degree E exceeding the secondary degradation judgment value TE2 in the judgment of the sequence A37, and the event described with reference to FIG. 7 has not occurred. When it is determined that the secondary failure countermeasure is not taken, the process returns to the process from the sequence A33, the operation from the process requesting information from the monitoring target blade module 200 is repeated, and the deterioration progress E is the secondary deterioration determination value. When the secondary deterioration determination condition is satisfied beyond TE2, or when at least one of the events described with reference to FIG. 7 has occurred, the state transits to a standby state in order to take measures against the secondary failure (sequence A39). ).

(15) The Manager 70 of the management server 100 terminates the use of the monitoring target blade module 200 being used by the user, and the Agent 10 of the monitoring target blade module 200 from the monitoring target blade module 200 in the above-described primary failure countermeasure standby state. When a power-off notification is received, in order to determine that the received power-off notification is not due to a restart or the like, the system continues to wait for a certain period of time, and if a power-on notification from the monitored blade module 200 is received during the standby No action is taken and the primary failure countermeasure standby state is continued until a power OFF notification is received again. If the manager 70 does not receive a power-on notification during standby, the manager 70 starts a process of switching the monitored blade module 200 to the spare blade module 210 as a countermeasure against the secondary failure. This switching process is performed by changing the computer name and IP address of the monitored blade module 200, the spare blade module 21, and 0 to switch the blade module to be operated to the spare blade module 210. This is processing for enabling the maintenance of the existing blade module as the monitoring target blade module 200 (sequence A38, A40 to A42).

(16) Before starting the switching process, the Manager 70 first communicates with the BMC 50 of the monitoring target blade module 200, issues a power ON command by IPMI, and performs a power ON operation of the monitoring target blade module 200 (sequence). A43-A44).

(17) Thereafter, when the Manager 70 receives a power ON notification of the blade module from the Agent 10 of the monitoring target blade module 200, the Manager 70 transmits a command to cancel the mirroring to both the blade modules 200 and 210, and mirrors to both the blade modules 200 and 210. Is terminated (sequences A45, A46, A48, A49).

(18) The Manager 70 of the management server 100 generates a batch file for rewriting the computer names of the blade modules 200 and 210 and the IP addresses of the OS and BMC, and sets a time difference between the blade modules 200 and 210. To send. This transmission is performed quickly to the blade module 200. The content of the process described in the generated batch file is to change its own computer name, rewrite its own OS IP address, rewrite its own BMC IP address, and reset the BMC (sequence). A47, A50).

(19) The Agent 10 of the monitoring target blade module 200 that has received the batch file from the Manager 70 of the management server 100 changes the computer name of the monitoring target blade module 200 to a dummy computer name according to the contents described in the file. A process of changing the IP address of the OS 20 and the BMC 50 to a dummy IP address is performed, and the BMC 50 is reset. Note that the dummy computer name and the dummy IP address are reserved in advance so that the number of spare blades existing in the system is equal to or greater than the number of spare blades (sequences A51 to A54).

(20) The Agent 11 of the spare blade module 210 that received the batch file a little later than the blade module 200 from the Manager 70 of the management server 100 sets the computer name of the spare blade module 210 of the monitored blade 200 according to the contents described in the file. The original computer name is changed, the IP address of the OS 21 and the BMC 51 is changed to the original IP address assigned to the monitoring target blade module 200, and the BMC 51 is reset (sequence A55 to A58).

(21) Thereafter, the spare blade module 210 that has become the monitoring target blade module notifies the manager 70 of the management server 100 that the change of the computer name and the IP address has been completed (sequence A59).

(22) The Manager 70 of the management server 100 that has received the notification that the change of the computer name and the IP address has been completed from the spare blade module 210 communicates with the BMC 50 of the blade module 200 that has been the monitoring target blade module so far. Then, the BMC 50 is instructed to perform the power OFF operation by IPMI, the BMC 50 of the blade module 200 is turned off, and the processing here ends (sequences A60 to A62).

  With the processing described above, the blade module failure prediction / switching processing is completed.

  FIG. 8 and FIG. 9 are sequence charts for explaining another example of processing for determining deterioration of the monitored blade module 200 and taking countermeasures against the failure in the client server system according to the embodiment of the present invention. To do. The processing of the sequence shown in FIGS. 8 and 9 to be described here is an example of processing in which the failure countermeasure from the normal operation is performed only once, and is a series of processing, so that FIG. 8 and FIG. 9 are continuous. It explains as being. In addition, the other example of the process described with reference to FIG. 8 and FIG. 9 is the same condition as the secondary deterioration determination condition in the example described with reference to FIG. 4 and FIG. Is equal to or greater than the secondary deterioration determination value TE2, or the event shown in FIG. 7 occurs.

  In the sequences shown in FIGS. 8 and 9, the processing operations in the sequences B1 to B11 from the monitoring of the blade module to the start of the failure countermeasure are the same as the processing operations in the sequences A1 to A11 in the example described with reference to FIGS. It is.

  Thereafter, as specific contents of the countermeasure against failure, first, the data of the monitoring target blade module 200 is copied to the spare blade module 210 (sequence B12 to B30). This process is the same as the data duplication process in the processes of sequences A12 to A27 and A29 to A31 in the primary failure countermeasure in the example described with reference to FIGS.

  In the sequence processing shown in FIGS. 8 and 9, after the data replication, the blade module to be operated is switched instead of mirroring (sequence B31 to B41). This process is the same operation as the blade module switching process as a countermeasure against secondary failure in the processes of sequences A47 and A50 to A59 described with reference to FIGS.

  With the processing described above, the blade module failure prediction / switching processing is completed.

  In other examples of the processing described with reference to FIGS. 8 and 9, the deterioration determination condition is the same as the secondary deterioration determination condition in the example described with reference to FIGS. 4 and 5, that is, the value of the deterioration progress E is 2. It is assumed that the next deterioration judgment value TE2 or higher or the event shown in FIG. 7 has occurred, but the deterioration progress level E may be higher than TE3 (TE2> TE3> TE1). Further, the event shown in FIG. 7 may be different from the example shown in FIG. 7, or only when the value of the deterioration progress E is equal to or higher than TE3.

  According to the above-described embodiment of the present invention, when it is determined that there is a high risk of failure of a device to be maintained such as an information processing device, it is possible to take measures to avoid the failure. In addition to reducing the time and effort of the administrator, it is possible to avoid a situation in which the blade module is damaged even though the risk of failure is predicted.

  In addition, according to the embodiment of the present invention, mirroring is performed only for a device that is determined to be at risk of failure, so the number of necessary devices is reduced compared to a normal redundant configuration. It is possible to reduce installation space, management effort, TCO, and the like.

  Further, according to the embodiment of the present invention, in normal failure prediction, there may be a case where an actual failure occurs earlier than the predicted failure date and time, but since mirroring as a countermeasure against failure can be performed early, The damage caused by the failure that is expected earlier can be suppressed.

  In normal failure prediction, when measures such as device replacement are taken early, the risk of device failure is low and safe, but the frequency of equipment maintenance increases and costs increase. On the other hand, when countermeasures such as replacement of devices are performed late, the frequency of device maintenance work decreases, but the risk of device failure occurring before the predicted failure date and time increases. According to the embodiment of the present invention, two steps of failure countermeasures are performed, mirroring is performed slightly earlier as the first step, and the device is replaced as a second step of failure countermeasures. Can be eliminated.

  Each process in the above-described embodiment of the present invention is configured by a program and can be executed by a CPU included in the present invention. These programs are stored in a recording medium such as an FD, CDROM, or DVD. It can be provided and can be provided by digital information via a network.

10, 11 Agent
20, 21 OS
30, 31 HDD (SSD)
40, 41 RAM with ECC
45, 46 Module body 50, 51 BMC
60-62 NIC
70 Manager
75 Control unit 100 Management server 200, 210 Blade module 300 Server 400 Client terminal 500 Network 600 Administrator terminal

Claims (8)

In the device failure prediction and countermeasure method that predicts the failure of the device to be maintained in advance and implements countermeasures against the failure,
A management server connected to the maintenance target device via a network;
The management server has management means for acquiring information on the device from the device to be maintained at a certain period and managing the device,
The management means calculates the deterioration progress of the device to be maintained based on the acquired information of the device to be maintained, and when the calculated deterioration progress is a predetermined progress, Alternatively, when the information acquired from the device to be maintained indicates a predetermined event, a failure prediction / measure method for implementing a failure countermeasure for the device to be maintained.   2. The failure according to claim 1, wherein the failure countermeasure for the device to be maintained is a measure for switching and operating the device to be maintained to a spare device different from the device to be maintained. Prediction and countermeasure methods.   The information collected from the maintenance target device by the management means is the S.P. M.M. A. R. T.A. The information includes hardware monitoring log, ECC error log of RAM with ECC, voltage value / temperature value of board, and cumulative number of forced power-off / forced reset performed on blade module. Item 1 or 2 failure prediction and countermeasure method. In the device failure prediction and countermeasure method that predicts the failure of the device to be maintained in advance and implements countermeasures against the failure,
A management server connected to the maintenance target device via a network;
The management server has management means for acquiring information on the device from the device to be maintained at a certain period and managing the device,
The management means calculates a deterioration progress degree of the maintenance target apparatus based on the acquired information of the maintenance target apparatus, and the calculated deterioration progress degree becomes a predetermined first progress degree. If the information acquired from the maintenance target device indicates a predetermined event, the first failure countermeasure is performed on the maintenance target device, and then the calculated degree of deterioration progresses. A failure prediction / countermeasure method characterized by implementing a second failure countermeasure for the device to be maintained when the second degree of advance is reached.   The first failure countermeasure for the maintenance target device is a countermeasure for mirroring and operating the maintenance target device and a spare device different from the maintenance target device. 5. The failure prediction / countermeasure according to claim 4, wherein the second countermeasure against failure is a countermeasure for switching the maintenance target device to a spare device different from the maintenance target device. Method.   The information collected from the maintenance target device by the management means is the S.P. M.M. A. R. T.A. The information includes hardware monitoring log, ECC error log of RAM with ECC, voltage value / temperature value of board, and cumulative number of forced power-off / forced reset performed on blade module. Item 6. The failure prediction / countermeasure method according to item 5 or 5. In the client server system that can predict the failure of the device to be maintained in advance and implement countermeasures against the failure,
A management server connected to the maintenance target device via a network;
The management server has management means for acquiring information on the device from the device to be maintained at a certain period and managing the device,
The management means calculates the deterioration progress of the maintenance target device based on the acquired information of the maintenance target device, and the calculated deterioration progress becomes a predetermined progress degree. Or a means for taking a countermeasure against a failure of the maintenance target device when the information acquired from the maintenance target device indicates a predetermined event. In the client server system that can predict the failure of the device to be maintained in advance and implement countermeasures against the failure,
A management server connected to the maintenance target device via a network;
The management server has management means for acquiring information on the device from the device to be maintained at a certain period and managing the device,
The managing means calculates a deterioration progress degree of the maintenance target device based on the acquired information of the maintenance target device, and a first progress degree in which the calculated deterioration progress degree is predetermined. Or when the information acquired from the device to be maintained indicates a predetermined event, means for implementing the first failure countermeasure for the device to be maintained, and then calculated A client server system comprising: means for implementing a second countermeasure against a failure with respect to the apparatus to be maintained when the degree of deterioration has reached a predetermined second degree of progress.