JP2002099522A - Message transfer method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To particularly increase the transfer efficiency of a message transfer method which sends a message from a first computer to a second computer which are connected through a network, and to efficiently transfer every message including a normal message. SOLUTION: On the occasion of sending a message from the first computer to the second computer, the second computer is requested for learning of a new message. When transferring the same message, from that time forward, only a message ID of the message is transferred and the transfer of the message text is dispensed with. In addition, a message is divided into a fixed part and a variable part, so that a transfer of a message text that has only a different variable part is dispensed with.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for transferring a message between computers connected via a network,
In particular, it relates to a method for improving transfer performance.

[0002]

2. Description of the Related Art Conventionally, when an event that occurs in a plurality of computers connected via a network is monitored by a management computer, a message corresponding to the event is transferred from the monitored computer to the management computer. When there are a large number of monitored computers, the number of messages is large, so that the load on the management computer becomes enormous, and the processing cannot be completed. For this reason, a technique is generally known in which a monitored computer filters a message and transfers only important messages such as fault information to a management computer.

There is also a technique for suppressing transfer of the same message as disclosed in Japanese Patent Application Laid-Open No. H11-224242.

[0004]

For example, in the filtering method, although the occurrence of a failure such as a failure message can be recognized by the management computer, there is a problem that detailed information leading to the failure cannot be understood. In particular, the normal message before the failure may contain information up to the failure, and the management computer receives the failure message and re-reads the previous normal message to the monitored computer. Must go.

Further, in Japanese Patent Application Laid-Open No. H11-224242, there is a problem that a message library must be transferred to a management computer in advance, and a message that may occur in an application program must be predicted. .

The present invention has been made in view of the above-mentioned problems of the conventional method, and has as its object to provide a message transfer method capable of efficiently transferring all messages including a normal message in a message transfer method between computers. And It is another object of the present invention to provide a method that can be applied without previously predicting a message issued by an application program.

[0007]

In order to achieve the above object, the present invention distinguishes message types in a message transfer method for sending a message from a first computer connected to a network to a second computer. Means for transferring the message ID, the message learning request command, and the message from the first computer to the second computer; and transmitting the message together with the message ID at the second computer to which the message learning request command has been sent. A means for storing, a means for sending only the message ID by the first computer if the message has a message ID for which the message learning request command has already been sent, and a means for sending the message ID sent from the first computer to the second computer. By having a means to assemble a message from the stored message ID, For one type of message, characterized in that it eliminates the need for transfer of the second and subsequent message text.

Accordingly, it is not necessary to transfer all the messages that may be transferred to the second computer in advance, and the learning is performed each time, and the minimum required storage is sufficient, so that there is no waste. No special storage area is required, and message retrieval is fast.

A message learning request can be explicitly specified by a learning request command. Alternatively, a message ID and a message can be sent first without being specified by a command, and learning can be performed using the message ID and message as a trigger. Further, it is also possible to make learning after sending a set of a message ID and a message a plurality of predetermined times.

Further, the message is divided into a fixed part and a variable part, the fixed part is stored, and the variable part is transferred together with the message ID. The transfer is not required.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the contents of the invention of message processing.

This system comprises a monitoring server 100 and a distribution node 110 and a network 130 connecting the monitoring server and the distribution node.

When the distributed node sends a message corresponding to an event that has occurred on the distributed node to the monitoring server, the message to be transmitted is compressed. The monitoring server can monitor the event of the distributed node without increasing the load on the network 130.

A distributed node includes a server for executing an application, a file server, a printer server,
Network equipment.

The distributed node 110 includes an event transmitting unit 111 for transmitting an event to the monitoring server, and an event transmitting buffer 11 for transmitting a plurality of events collectively.
2. The time slot management unit 113 which requests the event transmission unit to transmit when the buffer is filled with the event, the buffer control unit 114 which monitors the storage state of the event, and the message text generated by the event matches the contents of the message library 116. A message comparing unit 115 for examining whether the message text is recorded, a message library 116 for recording a message text, an event receiving unit 117 for receiving an event occurring in the node, an API (Applic).
ation Program Interface) 11
8. AP (Application Program)
119, a log conversion unit 120 that converts logs of OS and AP generated in the node into events, and a log 121.

The AP 119 reports events such as the start / end of a job and occurrence of an error to the event receiving unit 117 using the API 118 in real time. The message notified by the event is divided into a fixed part indicating the content of the message and a variable part indicating the time or the job name.

The event receiving unit 117 passes the message to the message comparing unit 115, retrieves the message ID from the message library 116, and
D is determined. If it is a new message ID, it is recorded in the message library 116 and the event transmission buffer 11
2 stores a message learning request command and a message. If the message ID is already present, the message is stored in the event transmission buffer 112. Note that the message learning request command may be configured to learn from a message to be sent first without being explicitly specified.

The message ID is a character string that combines alphabets and numbers added to the beginning of message text displayed and output by the system or application to the operator or user. The message ID indicates the cause of the message output according to the output program and the degree of importance. If the cause is the same, the message text to be displayed and output is the same, and is unique within the operating system.

When an event is stored in the event transmission buffer, the buffer control unit 114 starts a timer of the time slot management unit. It is checked whether the event transmission buffer 112 is filled with the event and the event transmission unit 111
The event in the event transmission buffer to the monitoring server 10
0, and the timer of the time slot management unit is stopped.

When the buffer control unit 114 searches for a message in the event transmission buffer and finds the same message as the message to be stored, the buffer control unit 114 stores relative reference information such as "same as the nth previous message". By storing relative reference information of a smaller size than the message, the message in the buffer can be compressed.

The timer of the time slot management section expires when the event transmission buffer 112 is not filled with the event within a predetermined time, and requests the event transmission section 111 to transmit the event in the event transmission buffer to the monitoring server 100.

A plurality of events can be transmitted at one time by waiting for an event transmission in the event transmission buffer for a fixed time using the timer of the time slot management unit. By transmitting a plurality of events at once, it is possible to reduce the number of times the events are transmitted, thereby suppressing an increase in load due to event transmission on the network.

The monitoring server 100 includes a receiving AP 101 for receiving a message, and a display unit 10 for displaying a message.
2. It comprises an API 103, a message assembling unit 104, an event receiving buffer 105, a message learning unit 106, a message library 107, and an event receiving unit 108.

Upon receiving the message learning request command and the message from the distributed node, the event receiving unit 108
Stores an event from a distributed node and a message learning request command in the event reception buffer 105. If there is no learning request command and there is only a message, it is treated as if the learning request command was omitted.

When a message learning request command is received from the distributed node 110, the message learning unit 106 of the event receiving buffer records the message text stored after the command in the message library 107.

When the event receiving unit 108 receives an event from the distributed node 110, the message assembling unit 104 reproduces the compressed message according to the message attributes recorded in the message library. If the content of the received message specifies reference information indicating that "the same as the nth previous message", the event reception buffer 1
The message in 05 is taken out and reproduced.

The reproduced message is passed to the receiving AP 101 via the API 103. The display unit 102 displays a message.

FIG. 2 shows the format of a message passed by the API 103 and the API 118.

The message has a message ID 200,
Event severity 201, message text 202,
Further, it is a text which is composed of a variable section 203 whose content changes when a time, a job name or the like in the message text occurs, and which is displayed by the distributed node.

The message ID is a text serving as a key at the time of retrieval.

The event priority 201 is determined by the message I
It indicates the type of event and the longitude of the failure corresponding to D, and is a fixed value for the message ID.

The message text 202 is a text excluding the variable part of the message. Instead of the variable part, a control code such as "[* 1]" is provided to indicate the start position of the variable part.

The variable section 203 stores the message text 20
2 stores the content that changes when the event occurs, such as the time displayed on the variable section 2 and the job name of the distributed node where the event occurred.

FIG. 3 shows a flow of processing between the modules of the distributed node.

The event generated by the AP 300 is written in a log (304), and the occurrence of the event is reported to the API 300 (305). The API 301 issues a log conversion request to the log conversion unit 330 (306).

The log converter 330 converts the log written by the AP 300 into a message (307), and passes the message to the event receiver 302 (308). Upon receiving the message, the event receiving unit 302 passes the message to the message comparison unit 303 (309), and the message comparison unit 303 searches for a message on the message library 116 using the message ID in the message as a key (310). The search result of the message ID is determined (311). If the message ID is not in the message library, the message is added to the message library (312).

If the message ID has already been output, and if the message is new, the message comparison unit passes the message learning command and the message to the buffer control unit 314 (314).

FIG. 4 shows a flow of processing between the modules of the distributed node.

The message comparison unit 400 passes a command or a message to the buffer control unit 401 (40
4).

The buffer control unit 401 calculates the total size of the message and the command (405). The remaining capacity of the event transmission buffer is compared with the total size of the message and the command (406). When the remaining capacity size of the transmission buffer is smaller than the total size of the stored message and command, the message time cannot be stored in the buffer. An event transmission request is sent to the event transmission unit 403, and the event in the event transmission buffer is passed (407). The event in the event transmission buffer is cleared, the remaining capacity size of the event transmission buffer is reset (408), and a reset request of the buffer monitoring timer is made to the time slot management unit 402 (409).

When the remaining capacity size of the event transmission buffer exceeds the total size of the message and the command, the message is compressed and stored in the event transmission buffer (4).
10). Message compression stores relative reference information of a smaller size than the message when the same message is present in the event transmission buffer.

If the monitoring timer of the event transmission buffer has not been set (411), a timer setting request is issued to the time slot management unit 402 (412).

The event transmission buffer is monitored by transmitting an event if there is an event in the buffer even if the event is not satisfied after a certain period of time in order to prevent the buffer from being waited for a long time without being filled with the event. .

When the timer expires (413), the time slot management unit 403 requests the event transmission unit 403 to transmit the event (414), clears the event in the event transmission buffer, and clears the remaining capacity of the event transmission buffer. The size is reset (415).

When receiving the event transmission request, the event transmitting section 403 transmits the event to the monitoring server (41).
6).

FIG. 5 shows a flow of processing between modules of the monitoring server.

When the event receiving section 500 receives the event, the event is stored in the event receiving buffer 105 (506). Event 1 in the event reception buffer
If the command is a learning request command (507), the request is made to the message learning unit 501 for learning (508), and the message learning unit 501 adds the message following the command to the message library 107.

When the processing of the learning request command in the event receiving buffer 105 is completed, the message assembling section 502
(510). The message assembling unit 501 finds the same message from the relative position information in the buffer with respect to the reference information in the event receiving buffer 105 and reproduces the message.
(510). After playing the message, set the event to API
Using (503) (512), the event is notified to the reception AP 504 and the display unit 505 (511, 513).

The display unit 505 displays the received event on the monitoring server (514).

FIG. 6 shows a flowchart of message compression in the buffer control unit.

When a message is requested to be stored in the event transmission buffer (600), the buffer control unit searches for a message in the buffer (601). If there is exactly the same message in the buffer (602), relative reference information such as "same as nth previous message" is stored in the buffer (603). If the exact message is not in the buffer, store the message in the buffer.
(604).

Since the size of the relative reference information is smaller than the message, the load on the event transmission buffer and the network can be reduced.

FIG. 7 shows a flowchart of the message comparing section.

A learning request command is received from the distributed server to the event receiving unit (701). The message comparison unit stores the message in the buffer in the message library (702).

The message library contains the message I
D, event importance, and message text are stored.

FIG. 8 shows a flowchart of the time slice.

When the message comparison unit issues a command or message storage request (801), the timer is set if the buffer monitoring timer has not been set. The purpose of storing a plurality of messages in the buffer is to reduce the network load by transmitting a plurality of messages at once and reducing the number of transmissions. The buffer monitoring timer transmits the accumulated messages at regular intervals to prevent the messages from being left in the buffer. If the monitoring timer of the buffer has not been set (802), the monitoring timer is set (803).

The sum of the size of the command requested to be stored and the size of the message is calculated (804). If the total size is larger than the remaining buffer size (80
5) Request transmission of the message in the buffer (80)
6), reset the buffer monitoring timer (807).
The event transmission buffer is cleared (808), and the remaining capacity of the buffer is reset (809). The command and the message are stored in the buffer (810), and the buffer monitoring timer is set (8111).

When the monitoring timer expires (812), commands and messages in the buffer are transmitted regardless of the remaining capacity of the buffer (813).

FIG. 9 shows a flowchart of the message assembling section.

When the relative reference information is found by searching the event receiving buffer (900), the nth previous message in the buffer is found based on the "same as the nth previous message" of the reference information (901). The message is reproduced by copying the nth previous message (90
2).

[0063]

According to the present invention, message transfer between computers can be executed efficiently.

[Brief description of the drawings]

FIG. 1 is a diagram showing an example of the overall configuration of a system to which a message processing method according to the present invention is applied.

FIG. 2 is a format diagram of a message.

FIG. 3 is a diagram illustrating a flow of control of a distributed node.

FIG. 4 is a view for explaining the flow of control of a distributed node.

FIG. 5 is a diagram for explaining the control flow of the monitoring server.

FIG. 6 is a flowchart of compression of a buffer control unit.

FIG. 7 is a flowchart of learning of a buffer control unit.

FIG. 8 is a flowchart of a time slice of a time slot management unit.

FIG. 9 is a flowchart of extension of a message assembling unit.

FIG. 10 is a message library structure diagram.

[Explanation of symbols]

100: monitoring server, 101: receiving AP (Applic
ation Program), 102: display unit, 10
3. API (Application Program)
Interface), 104 ... message assembly part, 1
05 ... event reception buffer, 105 ... event reception buffer, 106 ... message learning unit, 107 ... message library, 108 ... event reception unit, 110 ... distributed node, 111 ... event transmission unit, 112 ... event transmission buffer, 113 ... time slot Management unit, 114 ...
Buffer control unit 115 message comparison unit 116
Message library, 117 ... event receiving unit, 11
8 API (Application Program)
Interface, 119 ... AP (Applica)
Tion Program), 120: Log conversion unit, 1
21: log, 130: network, 200: message ID of the message, 201: importance of an event corresponding to the message, 202: message text of the message, 203: message text part that changes each time a time or a job name is generated, 1000: message ID on the message library; 1001, message text on the message library; 1002: importance of the event corresponding to the message on the message library.

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Shoichiro Katahira 5030 Totsuka-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture F-term in the Software Division of Hitachi, Ltd. 5B045 BB31 BB47 BB48 BB49 JJ02 5B089 GA11 GB03 JA32 JB16 KA06 KA08 KB10 KC50 5K030 HA05 HB16 JA10 LE01 LE12

Claims (4)

[Claims] 1. A method of transferring a message from a first computer connected to a network to a second computer, the method comprising a message ID for distinguishing a type of message, and the message being transmitted together with the message ID to the first computer. Means for transferring from the computer to the second computer;
Means for storing the message together with the message ID sent by the second computer, and, if the message is already stored in the second computer, the message ID in the first computer
Means for sending only the message ID sent from the first computer to the message ID stored in the second computer
A message transfer method characterized by having means for assembling a message from a message. 2. A method for transferring a message from a first computer connected to a network to a second computer, the method comprising a message ID for distinguishing a message type, a message ID, a message learning request command, Means for transferring the message from the first computer to the second computer, means for storing the message together with the message ID in the second computer to which the message learning request command has been sent, and message ID for which the message learning request command has already been sent. If the message is the message of the first computer, means for sending only the message ID from the first computer and means for assembling the message ID sent from the first computer from the message ID stored in the second computer are required. Characterized message transfer method. 3. The method according to claim 1, wherein the message is stored when a message having the same message ID is received a predetermined number of times. 4. The message transfer method according to claim 1, wherein the message is divided into a fixed part and a variable part, the fixed part is stored, and the variable part is transferred together with the message ID.