JP3912673B2 - Network management apparatus and computer program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a network management device and a computer program suitable for transmitting large-capacity, real-time data such as an AV stream via a network.
[0002]
[Prior art]
In general, when data such as a stream is transmitted via a network, a stream transmission may fail due to a limitation of a buffer capacity of a switching device that configures the network and relays transmission data. Therefore, for such a case, it is advantageous if the number of switching devices is detected in advance at the transmission source, and the transmission method can be changed when a failure occurs.
[0003]
[Problems to be solved by the invention]
The present invention has been made in view of the above points, and an object of the present invention is to provide a network management device and a computer program that can detect the number of switching devices that constitute a network and relay transmission data.
[0004]
[Means for Solving the Problems]
In order to achieve the above object, the present invention transmits a plurality of transmission packets having different sizes to a destination via a network having a switch device, and reduces the delay time and size of each packet and the time required for transfer per unit amount. Based on this, the number of switch devices in the network is calculated.
That is, according to the present invention, means for transmitting a plurality of transmission packets of different sizes to a destination via a network having a switch device ;
Means for detecting a time at which each of the plurality of transmission packets has arrived at the destination;
When transmitting a plurality of transmission packets having different sizes, when the delay time from when all the packets are accumulated in the switch device to when the transfer is started is the same, each of the plurality of transmission packets The network delay time of each transmission packet is calculated from the transmission time and the detected arrival time, and the number of switch devices in the network based on the delay time and size of each transmission packet and the time required for transfer per unit amount Means for calculating
A network management apparatus is provided.
The present invention also provides a computer program for causing a computer to execute the above processing.
[0005]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a first embodiment of a network according to the present invention, FIG. 2 is a timing chart showing the relay timing of each switching hub when a relatively short packet is transmitted in the network of FIG. FIG. 4 is a timing chart showing the relay timing of each switching hub when a relatively long packet is transmitted in the network of FIG. 1, and FIG. 4 is an explanatory diagram showing a sequence for calculating the number of switching hubs.
[0006]
As an example, an Ethernet (R) network composed of switching hubs employs a transfer method called a store and forward method. The store-and-forward method is a method in which when a switching hub relays a packet, it stores (stores) the packet in an internal buffer while receiving the packet, and then transfers (forwards) it to the next device after accumulating the entire packet. .
[0007]
FIG. 1 shows an example in which three switching hubs # 1, # 2, and # 3 exist between the transmission source host A and the transmission destination host B. FIG. 2 shows the transmission source host A-switching hub # 1 in FIG. Shows the transfer (relay) timing of the packet a between, the packet b between the switching hubs # 1 and # 2, the packet c between the switching hubs # 2 and # 3, and the packet d between the switching hub # 3 and the destination host B ing.
[0008]
The packet a transmitted from the transmission source host A to the switching hub # 1 is stored (stored) in a buffer in the switching hub # 1, and the time required to complete this storage is the Ethernet (R). This corresponds to the time required to complete the transfer of the packet in the absence of the packet. Transfer of the packets accumulated in the buffer in the switching hub # 1 is started to the next switching hub # 2 after a certain delay time α after all the packets are accumulated. Thus, every time it passes through the switching hub, the time required for packet transfer increases. Furthermore, the delay time α may vary depending on the internal configuration of the switching hubs # 1 to # 3, and may vary depending on the network conditions.
[0009]
FIG. 3 shows the timing when a packet having a larger size than that shown in FIG. 2 is transmitted. The delay time α from the completion of packet accumulation to the switching hub # 1 until the start of packet transfer is shown in FIG. However, the time required from the start to the end of packet accumulation varies depending on the packet size.
[0010]
As can be seen from this, when two types of packets having different sizes are transmitted or transmitted / received between the hosts A and B, the total transfer time varies depending on the number of switching hubs and the packet size. Therefore, the difference in the packet transfer time is measured, and the number of switching hubs # 1 to # 3 between the hosts A and B can be known based on the packet size.
[0011]
The first embodiment will be described with reference to FIG.
(1) First, an arrival time request packet of size S0 is transmitted from host A to host B.
(2) The host B that has received the arrival time request packet notifies the host A of the arrival start time as the arrival time of the packet.
(3) Host A calculates transfer time t0 from the transmission start time and arrival start time of the packet of size S0 (not shown).
(4) Next, an arrival time request packet of size S1 is transmitted from host A to host B,
(5) Host B that has received the arrival time request packet notifies host A of the arrival start time as the arrival time of the packet.
(6) Host A calculates transfer time t1 from the transmission start time and arrival start time of the packet of size S1 (not shown).
[0012]
Here, when a packet having a size of s [bytes] is transmitted on a transmission path having a time required for transfer per byte of T [s], the time required from the start of packet transfer to the end is s × T [S]
It is. Therefore, the time required from the start of packet arrival to one switching hub until the start of transfer of this packet is s × T + α [s]
In the case of n switching hubs, the time t from the start of transfer of host A to the start of arrival of host B is t = n × s × T + Σα (1)
It becomes. Therefore, substituting time t0 and t1 when transferring packets of size S0 and S1 into equation (1),
t0 = n × S0 × T + Σα
t1 = n × S1 × T + Σα
Therefore, Σα is deleted and n = (t0−t1) / {T (S0−S1)} (2)
It becomes.
[0013]
A second embodiment will be described with reference to FIG.
(1) First, the host A requests an arrival time from the host B, and transmits a packet of the same size that requests packet transmission of the same size and notification of the transmission time,
(2) The host B that has received the packet notifies the host A of the arrival start time of the packet.
(3) Host A calculates transfer time t0 (a) from the transmission start time and arrival start time of the packet of size S0 (not shown).
(4) Next, a packet of the same size S0 is transmitted from host B to host A.
(5) Also notify the transmission start time of the packet.
(6) Then, the host A calculates the transfer time t0 (b) from the transmission start time and the arrival start time (not shown).
[0014]
(7) Next, the host A requests the arrival time from the host B, and transmits the same size packet transmission and the size S1 packet requesting the transmission time notification.
(8) The host B that has received the packet notifies the host A of the arrival start time of the packet.
(9) Host A calculates transfer time t1 (a) from the transmission start time and arrival start time of the packet of size S1 (not shown).
(10) Next, a packet of size S1 is transmitted from host B to host A,
(11) Also notify the transmission start time of the packet.
(12) Then, the host A calculates the transfer time t1 (b) from the transmission start time and the arrival start time (not shown).
(13) The host A further averages the transfer times t0 (a) and t0 (b) of the packet of size S0 to obtain the transfer time t0, and the transfer times t1 (a) and t1 (b of the packet of size S1. ) Is averaged to obtain the transfer time t1, thereby obtaining the number n from the equation (2) (not shown).
[0015]
Here, since the delay time α by the hub may vary depending on the network conditions, an error occurs in the calculation of the number n. Therefore, the two types of calculation packets having different sizes are transmitted or transmitted and received twice or more to investigate variation, or by measuring with three or more packet sizes, the number of packets can be measured even if the delay time α by the hub fluctuates. n can be detected more accurately.
[0016]
【The invention's effect】
As described above, according to the present invention, a plurality of packets having different sizes are transmitted to the other party via a network having a switch device , based on the delay time and size of each packet and the time required for transfer per unit amount. Since the number of network switch devices is calculated, the number of switching devices that configure the network and relay transmission data can be detected.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a first embodiment of a network according to the present invention.
FIG. 2 is a timing chart showing relay timing of each switching hub when a relatively short packet is transmitted in the network of FIG. 1;
FIG. 3 is a timing chart showing relay timing of each switching hub when a relatively long packet is transmitted in the network of FIG. 1;
FIG. 4 is an explanatory diagram showing a sequence for calculating the number of switching hubs in the first embodiment.
FIG. 5 is an explanatory diagram showing a sequence for calculating the number of switching hubs in the second embodiment.
[Explanation of symbols]
A, B Host # 1 to # 3 Switching hub

Claims (3)

Means for transmitting a plurality of transmission packets of different sizes to a destination via a network having a switch device ;
Means for detecting a time at which each of the plurality of transmission packets has arrived at the destination;
When transmitting a plurality of transmission packets having different sizes, when the delay time from when all the packets are accumulated in the switch device to when the transfer is started is the same, each of the plurality of transmission packets The network delay time of each transmission packet is calculated from the transmission time and the detected arrival time, and the number of switch devices in the network based on the delay time and size of each transmission packet and the time required for transfer per unit amount Means for calculating
Network management device provided. Means further comprising means for detecting the time at which the destination has sent a reply packet of the same size as the transmission packet;
The network delay time of each transmission packet is calculated from the transmission time of each of the plurality of transmission packets and the detected arrival time, and each reception from the detected transmission time and arrival time of each of the plurality of reply packets. Calculating the delay time of the network of the packets to average the delay time for each packet size, and calculating the number of switch devices of the network based on the averaged delay time and size of each packet The network management device according to claim 1. Transmitting a plurality of transmission packets of different sizes to a destination via a network having a switch device ;
When transmitting a plurality of transmission packets having different sizes, each of the plurality of transmission packets has the same delay time from when all the packets are accumulated in the switch device until the transfer is started. Detecting the time of arrival at the destination;
Based on the transmission time of each of the plurality of transmission packets and the detected arrival time, the network delay time of each transmission packet is calculated, and based on the delay time and size of each transmission packet and the time required for transfer per unit amount Calculating the number of switch devices of the network
Computer program provided.

Images