JP2000250833A - Operating information acquisition method in multiple server operation management and recording medium recording the program - Google Patents

Abstract

(57) [Summary] [Problem] To improve the efficiency and reliability of operation management by performing the same fine-grained management as in the case where an operation manager is assigned to each server, and by reducing the load on the network and the management server. Significant improvement. In a centralized management system of a plurality of servers,
The management item for specifying the type of log data is set in the
0, 30, 40, agent module 2
Self-confirmation is made at 0, 30, and 40, and only necessary information is transmitted from the communication mechanisms 23, 33, and 43 to the management server 10 by collation of the filtering mechanisms 22, 32, and 42. The management server 10 distributes the defined management items and their conditions to the managed servers 20, 30, and 40, acquires log data corresponding to the managed items, and stores the log data in the storage device 110 together with the target server identifier. .

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for centrally managing the operation of a plurality of server computers. In particular, the present invention relates to a method in which a management item is provided for each server and an operation manager is provided for each server. The present invention relates to a method for acquiring operation information in a multiple server operation management, and a recording medium on which the program is recorded.

[0002]

2. Description of the Related Art Recently, with the progress of downsizing and openness of information processing systems, the number of cases in which a corporate information system is constructed by a client / server type system (hereinafter abbreviated as C / S type system) has increased. ing. Along with this, operation management of distributed systems has become an important issue. On the other hand, software technology for centrally managing system operation management has already been put to practical use. Regarding this kind of conventional technology, for example, see “Information Processing Society of Japan 98-
DSM-10 "is known. This technique will be described below. The overall configuration of this system includes a plurality of server computers (hereinafter, abbreviated as servers) to be managed and a monitoring server connected to the plurality of servers via a communication line. The group of servers to be managed has a function of collecting monitoring data using a module called an agent, and a function of communicating with a manager described later. The management server communicates with agents using a module called a manager,
Has a function to aggregate data from agents. The agent collects data such as fault information and configuration information (for example, CPU type, I / O configuration, etc.) that has occurred in the server, and operation logging information, and transmits the data to the manager. In addition, the manager confirms whether the server is alive or dead in addition to the data from the agent by PIN.
Using the G command (Packet Internet Groper: a command used to send small packet data to the other computer over a TIP / IP network such as the Internet and use the return time to check the status of the other computer or communication line) Running at regular intervals.

As a conventional centralized management method for devices and the like, there is, for example, a network management system described in JP-A-9-167126. This is because when a failure occurs, the server can determine which client has failed and what is the cause of the failure.
In this method, an identifier of an application executed on a client on the N network is registered, the identifier is registered in each client, and a log is obtained. However, in this method, all information is collected in the management server, resulting in excessive traffic.

[0004]

(A) The conventional centralized monitoring system uses an OS (Operat) of an operating server.
ing System), AP (Application)
(Program) is aggregated in the management server using a network, but in order to perform detailed operation management, a large amount (all) of information is collected in the management server, resulting in excessive traffic. There has been a problem that the number of processing steps for information increases and a line failure or the like occurs. (B) In addition, although the PING command can confirm the physical life of the server, there is also a problem that the life of the serving AP cannot be confirmed. (C) Even if the service AP is alive, D used by the AP
There is also a problem that the managed server does not have a means for checking the free space of the file capacity such as B (Data Base). (D) When the managed servers operate in different tasks, management and monitoring are required for each task. Therefore, a management and monitoring PC / terminal is required for each task. Due to the large number and the distributed monitoring terminals, it is difficult to manage manually,
There are commercially available products that can be centrally managed, but it is not possible to centrally (integrally) manage different management items for each task, and there is a risk that fault detection may be delayed due to excessive traffic. There's a problem.

Accordingly, an object of the present invention is to solve the above-mentioned problems of the prior art, and to perform a fine-grained management of a plurality of servers to be managed as if an operation manager exists for each server. In addition, a method for acquiring operation information in a multi-server operation management capable of reducing the load on a network or a management server, collecting failure notification information, and improving the efficiency and reliability of operation management, and a recording medium storing the program are provided. To provide.

[0006]

In order to achieve the above-mentioned object, in the operation information acquisition method in the multiple server operation management according to the present invention, necessary data among log data managed for each of a plurality of managed servers is provided. Only to the management server efficiently. That is, first, a client / server system in which a plurality of client computers are connected to a server computer,
An operation information acquisition method in a multiple server operation management of a system including a management server computer connected to a server computer of a plurality of client / server systems, wherein the management server computer includes log data acquired by the server computer. Defining a management item for specifying the type of the management item and the condition of the management item in association with the identifier of the server computer, and the defined management item and the condition of the management item correspond from the management server computer. Distributing to each of the server computers, wherein the plurality of managed server computers acquire log data corresponding to the distributed management items when the server computers operate, The step of storing log data and the acquired Transmitting the acquired log data to the management server computer together with the identifier of its own server computer depending on whether or not the log data satisfies the condition of the management item.

Second, a storage medium storing a program for realizing a method for obtaining operation information in the operation management of a plurality of servers, wherein a management item for specifying a type of log data to be obtained by a server computer and the management item And a process of distributing the defined management items and the conditions of the management items from the management server computer to the corresponding server computers, respectively. Performed by a server computer, when the server computer is operating, a process of acquiring log data corresponding to the distributed management item, a process of storing the acquired log data,
A program that causes the server computer to transmit the acquired log data to the management server computer together with its own server computer identifier depending on whether the acquired log data satisfies or does not satisfy the condition of the management item. It is characterized by recording.

[0008]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram of a multiple server centralized management system according to an embodiment of the present invention. FIG.
In the figure, reference numeral 10 denotes a management server which connects to and manages a client / server type system; 20 to 40, servers (A), (B), (C) to be managed; and 50 to 70, management servers 10 And managed servers (A) 20, (B) 3
0, (C) 40. Management server 10
Has an external storage device 110, and similarly, the managed servers 20, 30, and 40 each have an external storage device 1
It has 20, 130, 140. Also, 80, 90
Is a client computer connected to the managed server 20. Although a plurality of client computers are also connected to the managed servers 30, 40, they are not shown here. The management server 10 registers management items
Distribution mechanism 11, received data storage mechanism 12, communication mechanism 13
have. The management item registration / distribution mechanism 11 stores the management target server name (for example, the management target server 20, the name server (A)) and the management items of the server (A) in the server (A).
In addition, a copy of the registration information is distributed to the server (A) of the managed server 20, which is registered in the external storage device 120 via the management server 20. Thereafter, the management server 10 stores the management items of the server (A) in its own external storage device 110 and also stores operation information such as failures transmitted from the managed servers 20 to 40 thereafter.

The managed servers 20 to 40 include monitoring information acquisition mechanisms 21, 31, 41, and a filtering mechanism 22,
32, 42 and communication mechanisms 23, 33, 43. The monitoring information acquisition mechanisms 21 to 41 are management items registered in the server (for example, (A)) (distributed from the management server 10), for example, an important AP name for confirming life and death for status confirmation, a response status At regular intervals, monitor whether the response from the AP for confirming is correct, the threshold of the file capacity used by the AP, the threshold of the performance maintenance, and the logging information issued by the OS and the AP. It has a function of passing information to the filtering mechanism (22 to 42). The filtering mechanisms 22 to 42 check whether or not the threshold value of the management item and the failure notification match the registered management item. Is exceeded, the request is sent to the management server 10 by requesting the communication mechanisms 23 to 43.
The external storage devices 120 to 140 are used for storing management items distributed to the server.

FIG. 3 is a diagram showing an example of a management item table distributed from the management server to the managed server. The table 25 shown in FIG. 3 shows the contents of the table distributed from the management server 10 to the managed server 20 and stored in the external storage device 120. At the same time, similar information is stored in the external storage device 110 of the management server 10 for each managed server. In the case of FIG. 3 described in the management item registration table, management items, ranges, management contents or thresholds, and remarks are described. As shown in the remarks column, the first prc 26 represents monitoring of a service process (service AP), and the range is
This is the number of service processes to be monitored (the number of service APs), and the management content column is filled with the names of the APs to be managed. For the second CPU 27, CP
A threshold of 80% is set as the setting of the usage rate of U. For the third Memory 28, a threshold of 90% is set as the setting of the memory usage rate. For the fourth Disc 29, the partition 1 has a usage rate of 85
%, And the partition 2 has a usage rate of 75%.

FIG. 2 is an operation flowchart of operation information acquisition according to an embodiment of the present invention. Here, in the managed server 20, the service AP (B) and the monitoring information acquisition mechanism 2
No. 1 shows the processing when it has already been started. The monitoring information acquisition mechanism 21 reads the distributed management items of the server 20 or reads the management items from the external storage device 120, and internally develops them as a table (step 101). A command for confirming the life or death of the service AP (B) is issued (step 102). This command
This is a command by which the monitoring information acquisition mechanism 21 causes the OS to list the APs currently operating, for example, “RegQuery”.
In the ValueEx function, the argument is Process. The result including the abnormality detection is received and the data and control are passed to the filtering mechanism 22 (step 103). Service AP
If (A) is operating, a pseudo command is issued and a response is waited for in order to confirm whether the service AP (A) is operating normally (step 104). For example, when the service AP (A) is a core business AP, this pseudo command causes the AP to periodically issue an event log,
It is a command that reads the event log. When the service AP (A) is an e-mail program, the command is a command for causing the administrator to issue an e-mail.

If there is no response within a certain period of time, or if the content of the response is incorrect, the result information and control are passed to the filtering mechanism 22 (step 105). The free space of the file used by the service AP (A)
Check whether the threshold is exceeded (Step 10)
6). The used files and thresholds are developed in the table 25 as management items. The result information and control, including the free space information of the file and the case where the threshold is exceeded, are passed to the filtering mechanism 22 (step 1).
07). Hereinafter, all the management items developed in the table 25 are similarly confirmed (step omitted). It is checked whether or not all the management items have been confirmed, and if all the management items have been completed, the process waits at certain intervals (step 1).
08). The monitoring information acquisition mechanism 11 repeats the above-described processing at predetermined time intervals. That is, re-execution is performed from step 102 by interruption at the end of the interval (step 109).

The filtering mechanism 22 records the passed information as logging information in the external storage device 120 (step 201), and determines whether the information is normal or abnormal. (Step 202). Data to be notified to the management server 10 only at the time of abnormality is created (step 203) and transmitted to the management server 10 (step 204). The management server 10 receives the information from the managed server 20 (step 30).
1). The received information is recorded in the external storage device 110 (step 302), and a notification such as a display on a monitor, a report by voice, a report by mail, etc. is made (step 303).

FIG. 4 is a diagram showing a data format transmitted to a management server according to an embodiment of the present invention. The data 35 transmitted from the communication mechanism 23 to the management server 10 at the request of the filtering mechanism 22 is, for example, information (abnormality information) in a format as shown in FIG. In the data format, remarks are described for each item. For example, as shown in FIG. 4, for the server type 36, the types of servers such as a mail server and a business server are described.
For the server name 37, a name for specifying the server among the server types is described. Also, management item event 38
Is displayed in Japanese for the management item, whether the service level exceeds the threshold,
An event such as a down event is described. As an example, "The application of the 10th mail server group Exchange server went down. Or "Mail server group E
The capacity of partition 2 desk of xchange server 10 has exceeded 76%. Is transmitted.

Each processing flow shown in the embodiment shown in FIG. 2 mainly shows the operation of a program executed on the managed server 20, but when implementing the present invention, a program for performing this processing is executed by the program. It is recorded on a recording medium such as a CD-ROM, and the CD-ROM drive of the managed server 20
In general, the program is read from a CD-ROM, installed on a disk of the managed server 20, and then executed. Furthermore, as a recent distribution form, a form in which a program is loaded from another computer connected to the disk of the managed server 20 via a network and executed is increasing, and the program of the present invention is loaded in such a form. Even if the program is installed in the managed server 20 and executed, the same effect as that of the present embodiment can be obtained. In any case, if the program is stored on a recording medium,
The present invention can be realized by executing the program at any place and at any time.

[0016]

As described above, according to the present invention,
In a centralized management system with multiple servers, management items are assigned to each server, and self-confirmation can be performed with the agent module, so that it is possible to perform detailed management as if there is an operation manager for each server, and it is necessary for the filtering mechanism Because only necessary information is sent to the management server,
Since the load on the network and management server can be reduced, and failure notification information can be easily collected.
The efficiency and reliability of operation management can be greatly improved.

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of an operation information acquisition system in a multiple server operation management according to an embodiment of the present invention.

FIG. 2 is an operation flowchart of information confirmation and collection showing one embodiment of the present invention.

FIG. 3 is a table format diagram of management item registration according to an embodiment of the present invention.

FIG. 4 is a diagram illustrating a data format transmitted to a management server according to an embodiment of the present invention.

[Explanation of symbols]

10 management server, 11 management item registration / distribution mechanism, 1
2 ... data storage mechanism, 13 ... communication mechanism, 21, 31, 4
1 ... monitoring mechanism acquisition mechanism, 22, 32, 42 ... filtering mechanism, 23, 33, 43 ... communication mechanism, 110, 12
0, 130, 140 ... external storage device, 50, 60, 70
... communication lines, 80, 90 ... client computers,
25: management item registration table, 26 to 29: management items,
35 ... data format to be transmitted to the management server, 36
-38: transmission items.

Continued on the front page F term (reference) 5B042 GA12 JJ03 MC40 5B045 BB02 BB03 BB12 BB28 BB49 JJ08 5B089 GA11 GB02 GB08 JA35 JB15 KA06 KA07 KA13 KB03 KC15 KC30 MC03 MC03

Claims (2)

[Claims] 1. Operation of a multiple server operation management system including a client / server system in which a plurality of client computers are connected to a server computer, and a management server computer connected to a server computer in the plurality of client / server systems. An information acquisition method, wherein the management server computer defines a management item for specifying a type of log data to be obtained by the server computer and a condition of the management item in association with an identifier of the server computer. Distributing the defined management items and the conditions of the management items from the management server computer to the corresponding server computers, respectively, wherein each of the server computers first operates when the server computer operates. Management items distributed Obtaining the relevant log data; storing the obtained log data; collating whether the acquired log data satisfies the condition of the management item; Transmitting the log data together with the identifier of its own server computer. 2. A recording medium on which a program for realizing an operation information acquisition method in a multiple server operation management is recorded, wherein the management server computer includes a management item for specifying a type of log data to be acquired by the server computer. A process of defining a condition of the management item in association with the identifier of the server computer; and a process of distributing the defined management item and the condition of the management item from the management server computer to the corresponding server computer. The server computer, during operation of the server computer, a process of acquiring log data corresponding to the distributed management item, a process of storing the acquired log data, and the acquired log data is Whether or not the condition is satisfied is checked, and the management server computer is checked against Recording medium for recording a program characterized by storing a program for performing processing of transmitting the result of the log data together with an identifier of its own server computer.