KR100450417B1 - Network device operation state management method and sytsem using multi-thread - Google Patents

Abstract

The present invention provides a network host state management method and system using multiple threads that can examine the operating state of a large amount of network equipment or host computer in a short unit time.

The network host state management method and system separately generates a sender of a ping request message and a receiver of a return message by using a thread structure, and the sender inquires state information of each network device or a host computer, thereby responding to an unresponsive network device or host computer. Repetitively transmits a ping request message, and the receiving unit is configured to record the status and response time of the network equipment or the host computer according to the received ICMP message.

As a result, the present invention has the effect of minimizing the operation state investigation time for a plurality of network equipment or a host computer, and providing compatibility that can be utilized without additional changes to existing network equipment.

Description

Network host state management method and system of multi-threaded structure {NETWORK DEVICE OPERATION STATE MANAGEMENT METHOD AND SYTSEM USING MULTI-THREAD}

The present invention relates to a network host state management method and system using multiple threads that can investigate the operation state of a plurality of network equipment in a short time period in a network environment operated by a large amount of network equipment and a host computer.

In a network environment, it is necessary to check whether a host that a user attempts to connect to is actually operating. For this purpose, an Internet program for checking a status of a counterpart host to be accessed on a network is provided. This is called a PING (Packet Internet Groper), which checks whether the address responds to a particular Internet address.

Such a ping program consists of a structure for inspecting the operation state of one network object to be investigated. Therefore, in order to investigate the operation state of a plurality of network equipment, it is necessary to ping the network equipment sequentially.

In addition, the ping program examines the received ICMP message by waiting for a minimum time after transmitting the ping message to the target network device to confirm the response from the target network device. Therefore, in order to investigate the operation state with respect to a large number of network equipment, it takes a response waiting time as much as the maximum number of network equipment to be investigated. Such long processing time causes delays in the irradiation time and is not suitable in an environment in which the operation state of a plurality of network equipments needs to be investigated in a short period of time.

That is, the conventional ping program is made for the purpose of simply investigating the operation of one target network device, and thus, there is a problem that it is difficult to use for periodic operation state verification of a plurality of network devices.

Therefore, a network management program such as a simple network management protocol (SNMP) has been used as a method for managing a network. Such a conventional network management program can provide detailed information about a target network device or a host computer, but when a large number of network devices are targeted, a large amount of network traffic is generated and a large amount of data must be processed. High performance is required to enable this.

By the way, the management of a large network is basically required to investigate the proper operation of the investigation target, and whether or not the proper operation of the network can be determined through the periodic operation state investigation of the management target.

Therefore, a management system for a large network should be able to perform a survey function for a large amount of network equipment in a short period of time, and to be able to perform a survey function with a simple structure in order to reduce network management costs.

However, as described above, the conventional ping processing program and the network management program do not satisfy this requirement, and thus there is a problem that it is not suitable for the management of large-scale network equipment.

SUMMARY OF THE INVENTION The present invention has been made to solve such a conventional problem and propose an appropriate method for managing large-scale network equipment. The purpose of the present invention is to eliminate ping response waiting time from a plurality of target network equipment and simultaneously to ping a plurality of network equipment. The present invention provides a method and system for managing network host status using multiple threads that can perform a large amount of network device status inspection in a short time period.

In addition, another object of the present invention is to provide a method and system for managing network host state using multiple threads, which enables periodic state inspection of large-scale network equipment without additional modifications to the target network equipment and the host computer.

1 is a basic conceptual diagram of a network to which the present invention is applied.

2 is an overall configuration diagram of a network host state management system of a multi-threaded structure according to the present invention.

3 is a diagram illustrating an exchange state of messages in a network host state management system having a multi-threaded structure according to the present invention.

4 is a flowchart illustrating a processing procedure of a network host state management method of a multi-threaded structure according to the present invention.

5 is a structural diagram of a message used for network host state management of a multi-threaded structure according to the present invention.

Explanation of symbols on the main parts of the drawings

101: network host state management system

110: analysis means

120 control means

130-1 to 130-n: Transmission means

140: receiving means

As a constituent means for achieving the above object of the present invention, the network host state management system using the multi-thread of the present invention, by analyzing the ping operation setting factors according to the ping processing request, the number of threads and the like ping operation conditions And an analysis means for setting a function and a receiving means for receiving response messages from the systems to be investigated and the response means for the number of threads to be generated, according to the operating conditions set by the analysis means. Main control means for controlling the repetition check of the transmission means by inspecting a plurality of transmission means, a plurality of transmission means generated by the main control means for transmitting a ping request message to each system to be investigated, and a ping request message sent from the transmission means. Receive and receive ICMP messages from multiple investigation target systems or routers In accordance with the information it characterized in that it comprises a receiving means indicating the operating state of each of the subject system.

In addition, in the network host state management system according to the present invention, the ping operation condition set by the analysis means includes the number of systems to be investigated to be examined for the operation state, the number of threads to be generated, the waiting time between the repeated surveys, , The minimum number of repetitions and the maximum number of repetitions.

In addition, in the network host state management system according to the present invention, the ICMP message received by the receiving means includes a ping response ICMP message transmitted from an investigation target system receiving the ping request message and a ping request message from the investigation target system. The number of unreachable ICMP messages from the router if forwarding is not possible, and the TTL over ICMP messages from the router if the arrival time to the system under investigation exceeds the time-to-live (TTL) time in the IP header. have.

In addition, the network host state management system according to the present invention includes an IP header including an IP address of a destination, a type field and a code field indicating a response state to a ping request, a checksum field for checking an error generated during transmission, ICMP message consisting of an identification field for identifying a target host, a serial number field, time data for calculating a packet round trip time, a serial number for identifying each system to be investigated, and a field for identifying a repetition survey number. And a ping message composed of arbitrary dummy data for satisfying the ICMP message standard.

In addition, the network host state management method using the multi-thread of the present invention for achieving the above object, the step of calling the ping, the number of the system to be investigated, the number of threads created and the repetition with reference to the list of the host to be investigated Setting a survey condition, generating a number of transmission means, one reception means, and a state value store of the investigation target host according to the set condition; and a plurality of transmission means generated to the target host. Transmitting a ping request message, receiving a return message for the ping request to the generated receiving means, changing a status value of the investigation target host according to the reception message, a status value of the investigation target host, and a repeated survey condition According to, determining whether or not to repeat the repeated survey for each of the plurality of target hosts, the repeated survey for each target host If it is necessary to repeat the investigation according to the determination result of the characterized in that the step consisting of repeating from the fourth step.

The present invention also provides a computer readable recording medium having recorded thereon a program for executing each of the above-described steps.

Accordingly, the present invention eliminates the waiting time according to the existing request-response structure through a multi-threaded structure for ping message transmission and ICMP message reception, thereby examining a large number of network equipment in a short time. The same message structure can be used to provide compatibility without requiring additional modifications for network management.

Hereinafter, the configuration and operation of a network host state management method and system using multiple threads of the present invention will be described with reference to the accompanying drawings.

1 is a diagram schematically showing a network environment to which the present invention is applied.

The network host state management system 101 of the multi-threaded structure according to the present invention transmits a ping request message to the target hosts 103 subject to the operation state investigation. The transmitted ping request message may be transmitted through one router 102 such as path 101a according to the location on the network of each of the target hosts 103 and through two or more routers 102 such as path 101b. It may be delivered or may be delivered directly to the target host 103 without passing through a router such as route 101c.

In the case where direct transmission is not performed via a router, the target host 103 exists on the same network as the ping processing system 101.

The target hosts 103 receiving the ping request message delivered as described above automatically return the ping response message to the ping processing system 101.

The router 102 returns an unreachable ICMP message to the ping processing system 101 if it cannot deliver the ping request message to the destination host 103 and exceeds the time-to-live (TTL) value set in the IP header. The TTL excess ICMP message is returned to the ping processing system 101 for the path.

2 is a processing structure diagram of a multi-threaded ping processing system in which transmission and reception of the ping request and response messages as described above are processed.

As shown, the ping processing system 101 is provided with an analysis target list and setting factors when a ping call is analyzed, and analyzes the setting means 110 to set the test conditions, and the test conditions set by the analysis means 110. The control means 120 generates a ping request message transmitting means and a return message receiving means corresponding to the number of threads according to the number of threads, and generates the number of threads when pinging the call by the control means 120. And a plurality of transmission means (130-1 to 130-n) for repeatedly sending a ping request message to a corresponding target host, and receiving a return message from a plurality of investigation target hosts or routers to receive ping processing results for each investigation target host. It includes a receiving means 140 for transmitting to the control means 120.

The operation of the multi-threaded network host state management system configured as described above is as follows.

The operational status survey of the equipment of the large network begins with the transmission of a call message including a file and a configuration parameter with a list of the hosts 103-1 to 103-n to be investigated to the system 101.

The file and the setting factor delivered together with the call of the system 100 are transferred to the analyzing means 110, and the analyzing means 110 analyzes the file having the passed factor value and the investigation target list to exchange the ping message. The number of threads to be created is set according to the recursive search condition and the number of objects to be examined. The operating conditions set by the analyzing means 110 are stored and shared in the position 'conditions'.

Then, the control means 120 generates a plurality of transmission means 130-1 to 130-n for transmitting a ping message, and a receiving means 140 thread for receiving an ICMP message, and also, Specify the save location to save the operation status. In this case, the designated storage location is referred to as shared target hosts 121.

The control means 120 generates the transmission means 130, which is the first one-time ping message transmitter, by the set number of threads n. The generated transmission means 130-1 to 130-n transmits a ping request message to the allocated investigation target hosts 103-1 to 103-n.

In addition, the ICMP message responded from the investigation target hosts 103-1 to 103-n is received by the receiving unit 140, and the ping message reception details of the target hosts are recorded in response to the ICMP message of the shared target hosts 121. Reflect to status value.

The control means 120 determines whether or not to repeat the irradiation according to the state value of the shared target hosts 121 and the repeated irradiation condition, and controls the plurality of transmission means (130-1 ~ 130-n). That is, a ping request message is repeatedly sent to the target host for which an appropriate return message has not been received or has reached an iteration period.

3 is a network structure diagram illustrating a structure in which a ping message is transmitted and types of ICMP messages in a network to which a multi-threaded network host state management system according to the present invention is applied.

In the above, the message 301a is a ping request message sent by the network management system 101 to the target hosts 103 which are the investigation target systems. The ICMP message that the system 101 receives for such a ping request message may vary depending on various factors.

First, in the case of the target host 103a existing on the same network as the system 101, the ICMP response message 303a from the target host 103a is composed of a ping response message and returned.

Secondly, if the destination hosts 103b, 103d are in a different network than the system 101, the ping request message 301a from the system 101 is via the routers 102. When the ping request message 301a is properly delivered by the router 102 to the target hosts 103b and 103d, the target hosts 103b and 103d may send an ICMP message which is a ping response message to the ping processing system 301. (303a) is returned.

Third, the ping response message 303a may also be forwarded to the ping processing system 101 via the router 102. That is, when the target host 103c has an IP address that does not exist, an ICMP message 302a informing the ping request message 301a of the unreachable from the router 102 is returned to the ping processing system 101. In addition, if the number of routers 102 that can reach but pass is greater than the TTL value of the IP header, a TTL over ICMP message 302b is returned from the router 102 to the ping processing system 101.

4 is a flowchart illustrating a processing procedure of a network host state management method of a multi-threaded structure according to the present invention.

The processing according to the present invention is executed as a ping call is generated (401).

At this time, the data for the ping processing, the list of objects to be examined, and the ping processing setting factors are transmitted along with the ping call.

Therefore, when the ping is called, the data is first analyzed to determine the number of systems 103 to be examined for operation status, the number of threads to be created, the waiting time between repeated surveys, the minimum number of repetitions, and the maximum number of repetitions. Ping processing operation condition information such as the above is set (402).

Next, a predetermined storage location for sharing the survey result, that is, the state information for the survey target system 103 is designated (403), receiving means for receiving an ICMP message according to the operating condition, and a ping request message. Generate as many transmission means as the number of threads for transmitting (404).

The generated transmitting means transmits the ping request message to the investigation target host allocated to each thread by referring to the shared state information of the target hosts (405).

The ping request message transmitted in this way is delivered to the target destination host via a router or the like along each path, and an appropriate ICMP message is returned from the destination host or the passthru router according to the delivery status. At this time, the returned ICMP message is a response message to the ping request, an unreachable notification message, or a TTL exceeded message.

The ICMP message is received by the generated receiving means (406), and the sharing state information is changed (407) according to the contents of the received ICMP message.

In addition, with reference to the change history of the shared state information and the set repetition irradiation condition for each repetition period, it is determined whether the repetition irradiation is performed (408, 409). When the determination result, the repetition irradiation is necessary, the step 405 is repeated, Repeat the ping request message.

5 is a structural diagram of an ICMP message used in a multi-threaded network host state management system according to the present invention.

As shown, ICMP messages delivered between routers 102 or target hosts 103 for network management may use the same ICMP message structure as existing pings.

That is, the ping message sent to the investigation target host 103 and returned is an IPMP message format 501 of 20 bytes, an ICMP message 502 for 8 bytes of ping, and an 8 byte packet round trip time in ICMP message format. Satisfies the time data 503 for calculating the data, the serial number 504 for identifying each of the two-byte investigation target systems, the one-byte field 505 for identifying the repeated investigation number, and the ICMP message standard. 9 bytes of random dummy data 506 to be used.

The ICMP message 502 includes a type field 502a for an 8-bit ping request, an 8-bit code field 502b, a 16-bit checksum field 502c, and a 16-bit identification field 502d. ) And a 16-bit serial number field 502e.

As described above, the method and system for managing a network host state of a multi-threaded structure according to the present invention eliminates the waiting time consumed in the existing ping process, thereby examining the operating state of a large amount of network equipment or a host computer at a unit time. As a result, it is possible to check the operation status in a short period. As a result, it is possible to investigate the operation of a large amount of network equipment or a host computer in a short period in a large network, and by using the same protocol as the existing ping processing. There is an effect that it is applicable without making additional modifications to the network equipment or host computer to be investigated.

Claims (5)

Analysis means for analyzing the ping operation setting elements according to the ping processing request and setting ping operation conditions such as the number of objects to be examined; According to the operating conditions set by the analyzing means, generating as many transmission means as the number of threads to be generated, and receiving means for receiving response messages from the systems under investigation, and examining the response results of the systems under investigation, Main control means for controlling the repetition test; A plurality of transmission means generated by the main control means for transmitting a ping request message to each system to be investigated; And And receiving means for receiving ICMP messages from a plurality of investigation target systems or routers receiving ping request messages sent from the transmission means and notifying the operation status of each investigation target system according to the contents of the received message. Network host state management system using threads. The method of claim 1, wherein the analysis means The number of systems under investigation that need to be investigated, How many threads to create, Wait time between repetitions, The minimum number of iterations, A network host state management system using multiple threads, wherein the maximum number of iterations is set. The method of claim 1, wherein the ping request message and the ICMP message transmitted and received in the system An IP header containing the IP address of the destination, An ICMP message comprising a type field and a code field indicating a response state to a ping request, a checksum field for checking an error occurring during transmission, an identification field for identifying a target host, a serial number field, Time data for calculating packet round trip time, A serial number to identify each system under investigation, A field for identifying a repeat survey number, A network host state management system using multiple threads, characterized in that it consists of arbitrary dummy data to satisfy the ICMP message specification. A calling step of calling a ping; A condition setting step of setting the number of systems to be investigated, the number of threads created, and the recursive irradiation condition by referring to the list of the investigation target hosts; A thread generation step of generating a state value storage of the number of transmission means, one reception means, and the investigation target host according to the set condition; A transmitting step of transmitting a ping request message to a target host by the generated plurality of transmitting means; Receiving step of receiving a return message for the ping request to the generated receiving means to change the state value of the investigation target host in accordance with the received message; A determination step of determining whether or not to repeat the irradiation of each of the plurality of investigation target hosts according to the state value of the investigation target host and the repetition irradiation condition; And And repeating the determination step from the transmission step to a set number of repetitions if the repetition check is necessary according to the determination result of the repetition check for each target host. A computer-readable recording medium having recorded thereon a program for executing each step according to claim 4.