JP2003258939A - Duplex monitoring method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a duplex monitoring method which can notify a user of the possibility of a setting error in duplexing in its early stage and lighten the load of network management operation by detecting mismatch in the duplex state of a communication port of LAN equipment and notifying the user of it. <P>SOLUTION: The error occurrence state of a frame sent and received by a MAC layer module 3 is monitored, on the occurrence of the error, a software driver 5 decides that the duplex is in mismatch when the attributes of its own port are full duplex, but measures a traffic quantity when the attributes of its own port are semi-duplex to decide that the duplex state is in mismatch only when the traffic is small, so that an LED driver 6 and an LED7 give warning notice. <P>COPYRIGHT: (C)2003,JPO

Description

Description: BACKGROUND OF THE INVENTION [0001] 1. Field of the Invention [0002] The present invention relates to a duplex monitoring method for a communication device connected to a LAN line, and more particularly to a method for detecting a mismatch of a duplex setting with a link partner. The present invention relates to a duplex monitoring method that can detect deterioration of communication performance in a network at an early stage. [0002] In recent years, a typical local area network (Local Area Network: LAN) uses an Ethernet (R) cable, and the Ethernet (R) is 10BAS using the commonly known yellow cable
In addition to E-5, three types of cables are used: 10BASE-2, which uses a thin coaxial cable, 10BASE-T, which uses a twisted pair cable, and 100BASE-TX. In each case, the signal flowing on the cable is the same, and it is possible to use a combination of cables. For example, the trunk system is wired with 10BASE-5, and each department is wired with 10BASE-T and 100BASE-TX. [0003] 10BASE-T is a connection system that realizes a transmission speed of 10 Mbps using a twisted pair cable.
TX is 100Mbps using twisted pair cable
Is a connection method that realizes In 10BASE-T and 100BASE-TX that use twisted pair cables, the cable connectors and segment lengths are the same, so even when migrating a 10BASE-T network to 100BASE-TX, existing cables can be used as they are. . In other words, whether to realize a 10BASE-T network or 100BASE-TX using the same twisted pair cable is determined by the specifications of the communication ports of the two LAN communication devices connected to both ends of the twisted pair cable. Things. [0004] Twisted pair cables 10BASE-T, 100BASE-
Specific examples of devices that constitute a LAN environment using TX include a repeater hub (RepeaterHUB), a switching hub (SwitchingHUB), and a layer 3 switch (Layer3Switch).
And data terminal equipment (DTE) such as a host computer, a personal computer, and a printer. The repeater hub is simply called a hub, and transmits a packet sent from a certain terminal to terminals connected to all ports (nodes), and each terminal selects data. Switching hubs are connected to Ethernet (R) devices (Ethernet
(R) card) physical address (MAC address)
Analyzes the data sent from the terminal to detect the destination,
By transmitting data only to the destination terminal, data can be exchanged by directly connecting ports required for communication. Therefore, the load on the entire network is reduced, and the security is improved. In addition, since data is temporarily stored to analyze the destination, it can be used to connect networks with different speeds. Typical examples are 10Base-T of Ethernet (R).
And 100Base-TX, full-duplex and half-duplex networks of different speeds can be mixed. Layer 3 switches are specialized for the purpose of high-speed routing processing at the network layer (Layer 3). Generally, routers have functions such as filtering and multi-protocol support that have been removed to increase speed. This is a switch that realizes low cost of equipment. [0006] In the case of a data terminal device such as a personal computer or a printer, a NIC (Network Interface Card, Et) that is an expansion card for connecting to a private network (LAN).
hernet (R) card, LAN card, LAN board) into an expansion slot etc.
An N cable (twisted pair cable) is connected to the NIC and used. Normally, there are two link attributes of a communication port of a device constituting a LAN environment, that is, speed and duplex. Speed is 10Mbit / sec, 100M
At present, 100Mbit / sec Fast Ethernet (R) is mainly used except for trunk lines such as a backbone. Duplex includes half duplex and full duplex. FIG. 3A shows a combination of link partners, which are devices connected via a LAN cable.
As shown in the figure, a computer having an NIC (Host Computer in the figure) and a repeater hub (RepeaterH in the figure)
UB), half-duplex operation is normally performed.
As shown in the figure, in the case of a computer having an NIC and a switching hub (SwitchingHUB in the figure), or in FIG.
As shown in (c), in the case of a computer having an NIC and a layer 3 switch (Layer 3 Switch in the figure),
Normally, full-duplex operation is performed. FIG. 3 is an explanatory diagram showing a combination of link partners and a normal duplex operation. Regarding which of the two link attributes described above is adopted between the LAN communication ports of the link partner connected by one LAN cable, the link exchanged between the two LAN communication ports Determined by the pulse signal. Specifically, when the link pulse signal generating side is, for example, a communication port corresponding to 10BASE-T, an NLP (Normal Link Pulse
e) generate a signal called 100BASE-
In the case of a TX-compatible communication port, a signal called an idle signal is generated. In the case of a communication port that supports auto negotiation and has a function of determining an attribute by negotiation of both link partners, a FLP (Fast Lin
k Pulse). On the link pulse signal receiving side,
In the case of a communication port having a parallel detection function, when either the NLP or the idle signal is detected, the port is set to 10 Mbit / sec.
(Half duplex) or 100 Mbit / sec (half duplex). In the communication port corresponding to the auto negotiation, either the NLP or the idle signal is detected, and the port is set to 10 Mbit / sec.
(Half duplex) or 100 Mbit / sec (half duplex). When the FLP is detected, a negotiation sequence for setting a link with the link partner is started, and an optimum communication mode is selected from modes supported by both the own port and the link partner. . Regarding the details of these various link pulses and automatic negotiation, all are described in "ISO / IEC 8802-3".
It is specified in "IEEE Std 802.3". Here, when two communication ports of a link partner connected by one LAN cable both support a full-duplex function, specifically, a computer having an NIC and a switching hub, In the case of a computer having an NIC and a layer 3 switch, the conditions for setting these to operate at full-duplex are as follows. One is a case where both communication ports are set to be fixed at full duplex, and the other is a case where both communication ports are set to auto-negotiation. If the above conditions are not satisfied, for example, if one of the communication ports is set to full-duplex fixed and the other is set to auto-negotiation, the sequence for auto-negotiation is not exchanged. The communication port set to auto-negotiation has a half-duplex setting, and CSMA / C for realizing half-duplex is set.
D (Carrier Sense Multiple Access with Collision D
etection) function works. The CSMA / CD function in the half-duplex port is specifically the following four typical functions. (1) Carrier detection is performed at the time of frame transmission, and transmission is performed after confirming that the link partner has not transmitted the frame. (2) Monitor whether the carrier is lost during transmission. (3) When a collision during transmission is detected, retransmission is performed. (4) Transmission is suppressed during reception. Of the above four functions, the function (1), in particular, enables the half-duplex port to avoid unnecessary frame collision in the network segment. However, for example, if the link partner is set to full duplex, the link partner side may
Since there is no / CD function, there is a possibility that a full-duplex port transmits a frame while transmitting a frame from a half-duplex port. In this case, the half-duplex port side detects a collision as a result Will do. When a collision is detected on the half-duplex port side,
If a collision occurs within 512 bits of transmission, up to 16 retransmissions are performed at random intervals, frames with more than 16 retransmissions will be discarded, and collisions of 512 bits will be discarded. If it occurs after transmission, the transmission frame is discarded. In any case, the collision detection delays the transmission of the frame and, in some cases, discards the frame, thereby deteriorating the effective frame transfer rate (transfer throughput) of the line. With the recent spread of the Internet and LANs, the number of cases in which many users have to connect and set network devices has increased, while devices in which both link partners support full-duplex functions have been introduced. On the other hand, at present, the above-described conditions for setting both to operate at full duplex are not generally known.
As user psychology, for example, a network administrator
Since the user has knowledge of communication, to realize high-speed communication, all ports of the switching hub are set to fixed full-duplex, and users of computers connected to the hub do not have much knowledge of communication. In such a case, there is a high possibility that the default setting of the communication port of the NIC of the computer is left as the automatic negotiation. As a result, the switching hub side of the link partner is set to full-duplex, and the NIC side of the computer is set to half-duplex, which increases the probability of a duplex mismatch condition. Incidentally, as a prior art relating to the operation mode of the link partner in the LAN environment, there is a conventional technique disclosed in
Japanese Patent Laid-Open No. 10-303937 published on Jan. 13, "10Base
Method and Apparatus for Providing -T / 100Base-TX Link Guarantee "
(Applicant: Hewlett-Packard Company, Inventor: Timothy Wakeley, etc.). This conventional technology, 10Base-T / 100Base-TX device, 10Base-T half-duplex, 100Base-TX half-duplex, without manually selecting each operation mode of all 100Base-TX, all the link partners and link It is an algorithm that guarantees that However, conventional communication devices and NICs do not have a function of detecting a duplex state mismatch with a link partner.
Even if the communication performance has deteriorated, there is a problem that it cannot be immediately determined whether or not the cause is duplex mismatch. The present invention has been made in view of the above circumstances, and detects inconsistency in the duplex state of a communication port of a LAN device and notifies the user of this.
An object of the present invention is to provide a duplex monitoring method capable of notifying a user of a possibility of a duplex setting error at an early stage and reducing a load of a network management operation. SUMMARY OF THE INVENTION The present invention for solving the above-mentioned problems of the prior art is directed to a LAN connection device which performs transmission and reception by setting a full-duplex or half-duplex attribute to a LAN communication port. Monitoring the error occurrence state of transmitted / received frames, and when an error occurs, if the attribute of the own port is full duplex, judges that the duplex state does not match and issues a warning notification, and notifies the own port If the attribute is half-duplex, the amount of traffic is measured, and only when the traffic is low, it is determined that the duplex status does not match, and a warning notification is issued. It can notify and reduce the load of network management work. Embodiments of the present invention will be described with reference to the drawings. The function realizing means described below may be any circuit or device as long as the function can be realized, and some or all of the functions may be realized by software. is there. Further, the function realizing means may be realized by a plurality of circuits, or the plurality of function realizing means may be realized by a single circuit. The duplex monitoring method according to the present invention comprises:
Monitors the error occurrence status of transmitted / received frames, and when an error occurs, if the attribute of the local port is full duplex, judges that the duplex status does not match and issues a warning notification, and if the attribute of the local port is half duplex. If you measure traffic volume,
Since it is determined that the duplex state does not match only when the traffic is small, a warning is issued, the possibility of a duplex setting error is notified to the user at an early stage, and the load of network management work can be reduced. A configuration for realizing the duplex monitoring method according to the embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram of a configuration for implementing a duplex monitoring method according to an embodiment of the present invention. As shown in FIG. 1, a configuration for realizing the duplex monitoring method according to the present embodiment is a communication port 1, a physical layer module 2, and a MAC layer module 3, which are conventional configurations.
And a software driver 5, an LED driver 6, and an LED 7, which are characteristic parts of the present invention, in addition to the CPU 4. Each part of the present invention will be described. The communication port 1 is a port for connecting a LAN cable. The physical layer module 2 is an OSI that divides communication functions into seven layers.
(Open system interconnection) This is a functional module that realizes the physical layer (first layer) in the communication protocol model. The characteristics of cables such as Ethernet (R), the shape of pins, and the electrical interface of physical media And functions based on rules such as a basic data modulation method, and is usually provided by a hardware device. The MAC layer module 3 is a functional module that implements a lower sublayer (media access control: MAC layer) of the data link layer (second layer) in the OSI communication protocol model, and is a frame (data transmission / reception unit). It functions based on rules such as the transmission / reception method, frame format, and error detection method, and is usually provided by a hardware device. Realize CSMA / CD function in Ethernet (R),
A physical communication path with a communication partner is secured, an error of data flowing through the communication path is detected, and the CPU 4 is interrupted when an error is detected.
Note that the repeater hub does not have the MAC layer module 3. The CPU 4 includes the physical layer module 2 and the MA
Controls the C layer module 3. In the NIC of the computer, the M executed by the executed software.
A data transfer interface between the AC layer module 3 and the upper layer is provided. However, the present invention is not limited to a device such as a switching hub for forwarding frames. If an error occurs during frame transmission / reception, MA
An interrupt is generated from the C layer module 3 to the CPU 4 and instructs a software driver 5 to be described later to start a duplex monitoring process. The software driver 5 performs a duplex monitoring process in accordance with an instruction from the CPU 4. The duplex monitoring process is a process of collecting error information, determining whether there is a duplex mismatch according to the setting state of the own port, and issuing a warning notification when the duplex mismatch occurs. Here, a specific processing flow of the duplex monitoring processing in the software driver 5 will be described with reference to FIG. FIG. 2 is a flowchart illustrating an example of a flow of a duplex monitoring process for realizing the duplex monitoring method of the present invention. The duplex monitoring process in the software driver 5 is executed by the CPU 4
It is started by an instruction from the server, and first collects and totals error information (100). Then, it is determined whether the setting of the own port is full-duplex (102). If the setting is not full-duplex (No), that is, if it is half-duplex, both communication ports of the link partner are half-duplex. Even in the case of (1), if the line occupancy of the frame rises to a certain level or more due to an increase in traffic, a collision may occur.
4) It is determined whether the traffic condition is large (many) (10)
6) If there is not much (No), it is determined that there is a duplex mismatch (110), and a warning is notified (11).
2) End the duplex monitoring process. Here, the determination as to whether the traffic state is large (large) may be made by comparison with a preset threshold value, and the threshold value is empirically obtained. On the other hand, in the process 102, if the setting of the own port is full duplex (Yes), it is an error that cannot occur at first if both communication ports are set to full duplex. Therefore, it is considered that there is a duplex mismatch, so that the process jumps to step 110. On the other hand, if the traffic state is high (Yes) in the processing 106, the line occupancy of the frame is increased to a certain level or more due to the increase in traffic, and it is highly likely that a collision has occurred. The monitoring process ends. The LED 7 is an LED (Light Emitting Diode) for notifying the user of a duplex mismatch. The LED driver 6 is a driver that performs control to turn on the LED 7 according to an instruction from the software driver 5. Note that the LED driver 6 and the LED 7 are warning notifying means for warning the user of the mismatch of the duplex.
Not only the LED lighting but also a warning sound such as a buzzer may be performed. Next, the operation for realizing the duplex monitoring method of the present invention will be described. The operation for realizing the duplex monitoring method of the present invention is performed, for example, when the own port is set to full duplex and the link partner is set to half duplex, and when the link partner transmits from the own port during transmission. The frame transmission may be interrupted by detecting a collision on the link partner side, and an alignment error or a CRC error of the received frame is detected by the MAC layer module 3 on the own port side, and an interrupt is generated in the CPU 4. As a result, the duplex monitoring process of the software driver 5 is started. Since the setting of the own port is full duplex, it is determined that the alignment error and the CRC error of the received frame are caused by the mismatch of the duplex, and the LED 7 is turned on as a warning notifying means. If the port is set to half-duplex and the link partner is set to full-duplex, and if the link is transmitted from the link partner during transmission, the MAC layer module 3 of the port itself , A collision is detected, errors such as discard of a transmission frame due to excessive collision (over 16 retransmissions), late collision (collision occurring after 512-bit transmission), and the like are detected.
An interrupt occurs in the PU 4 and the duplex monitoring process of the software driver 5 is started. Since the setting of the own port is half-duplex, even if both communication ports are half-duplex, a collision may occur if the line occupancy of the frame rises to a certain level or more due to an increase in traffic. Software driver 5 measures the current traffic state, and it is small (low) that no error occurs during normal operation
When the occurrence of a high-frequency error is detected despite the traffic level, it is determined that the error is caused by the mismatch of the duplex, and the LED 7 is turned on as a warning notification unit. In addition, as a warning notification unit when duplex mismatch is detected, not only LED lighting but also a warning sound such as a buzzer may be used. According to the duplex monitoring method according to the embodiment of the present invention, when an error is detected in the MAC layer module 3, the setting of the own port by the software driver 5 is full duplex or half duplex. In the case of full-duplex, a warning is issued when there is a duplex mismatch, so the possibility of a setting error is detected early and reported to the user to reduce the burden of network management work. There is an effect that can be done. When the setting of the local port is half-duplex, even if both communication ports are half-duplex, a collision may occur if the line occupancy of the frame becomes higher than a certain value due to an increase in traffic. Therefore, the software driver 5 measures the current traffic state and detects a low (low) traffic level that does not cause an error during normal operation, but detects a high-frequency error occurrence. Since the warning is notified by determining that the setting error is caused, the possibility of the setting error is found at an early stage and the user is notified, and the load of the network management work can be reduced. According to the present invention, when it is determined that the duplexes do not match, the user is notified by a notification method such as LED lighting or buzzer alarm sound, so that the notification can be made easily and easily using a simple and inexpensive configuration. . According to the present invention, an error occurrence state of a transmitted / received frame is monitored, and when an error occurs, if the attribute of the own port is full duplex, it is determined that the duplex state does not match. If the attribute of the local port is half-duplex, a warning is issued and the amount of traffic is measured. Only when the traffic is low, the duplex status is judged to be inconsistent. There is an effect that the possibility of the setting error is notified to the user at an early stage and the load of the network management work can be reduced.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a configuration for realizing a duplex monitoring method according to an embodiment of the present invention. FIG. 2 is a flowchart illustrating a flow example of a duplex monitoring process for realizing the duplex monitoring method of the present invention. FIG. 3 is an explanatory diagram showing a combination of link partners and a normal duplex operation. [Description of symbols] 1 ... communication port, 2 ... physical layer module, 3 ... MA
C layer module, 4 CPU, 5 software driver, 6 LED driver, 7 LED

Claims (1)

Claims 1. A duplex monitoring method for a LAN connection device that performs transmission and reception with a full-duplex or half-duplex attribute set to a LAN communication port, wherein an error occurs in a frame transmitted and received. When an error occurs, if the attribute of the local port is full duplex, it is determined that the duplex status does not match and a warning is notified.If the attribute of the local port is half duplex, the traffic volume is measured. ,
A duplex monitoring method, comprising: determining that a duplex state does not match only when traffic is low; and issuing a warning notification.