JP2002033733A - Host computer and network system - Google Patents

Abstract

(57) [Summary] The present invention relates to a host computer, and it is an object of the present invention to measure the processing speed of file transfer between host computers connected to a network relatively reliably. A transmitting unit that transmits a predetermined measurement packet for measuring a network distance to a plurality of host computers connected to a network, a receiving unit that receives a reply packet to the measurement packet, and a reply packet A distance measuring unit that calculates a network distance to each of the host computers by using, a time measuring unit that measures a reply time from when the measurement packet is transmitted to when the reply packet is received, and for each of the host computers. And a determination unit for classifying a plurality of host computers using the network distance and the response time measured.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a host computer, and more particularly, to a host computer connected to various networks such as the Internet.

[0002]

2. Description of the Related Art When a host computer (hereinafter referred to as a host) reads a file from another host via a connected network, if a plurality of hosts can be selected, a small packet is conventionally transmitted and received between the hosts. Then, only the time required for the transmission and reception is measured, and a host having a high communication speed is selected.
Japanese Patent Laid-Open No. 7-20 / 1994 discloses a broadcast communication method for improving the reliability of broadcast communication using a network distance.
No. 0505. This is to statically check the network structure in advance, and cannot be used in a network such as the Internet where the network structure cannot be grasped in advance.

[0003]

When file transfer is performed using a network whose network configuration cannot be predicted in advance, such as the Internet, the communication speed between two hosts changes every moment. In the case of transfer or the like, the communication speed may greatly change before and after the transfer is started, and it is more reliable to judge the transfer speed only from the time actually used for the transmission and reception.

When there are a plurality of routes between two hosts, generally, as the number of other hosts existing on the route between the two hosts increases, the possibility of a change in communication speed increases. For this reason, when selecting a host with a high communication speed, it is necessary to consider the number of hosts existing on the route in order to increase the reliability of the selection.

Accordingly, the present invention has been made in view of the above circumstances, and considers the number of other host computers in a path between two host computers, thereby enabling a plurality of host computers. An object of the present invention is to provide a host computer connected to a network which can select a host computer which can perform processing such as file transfer quickly with high reliability.

[0006]

SUMMARY OF THE INVENTION The present invention provides a transmitting unit for transmitting a predetermined measurement packet for measuring a network distance to a plurality of host computers connected to a network, and a reply packet for the measurement packet. Receiving unit, a distance measuring unit for determining a network distance to each of the host computers using a reply packet, and a time for measuring a reply time from transmitting the measurement packet to receiving the reply packet. It is an object of the present invention to provide a host computer, comprising: a measuring unit; and a judging unit that classifies a plurality of host computers using the network distance and the response time measured for each of the host computers.

[0007] After the determination unit classifies the plurality of host computers according to the magnitude of the network distance, the host computer belonging to the same network distance N is returned to the smallest one of the host computers belonging to the network distance N + 1. It is an object of the present invention to provide a host computer characterized by classifying the host computers belonging to the network distance N into two, using a response time of the host computer whose time is measured as a threshold value.

[0008] Further, an HTML analysis processing unit for extracting hyperlink information from the received data, a site extraction processing unit for creating a list of connection destination host computers based on the extracted hyperlink information, The display device may further include a display unit that performs classification display of the host computers in accordance with a result of the classification of the host computers in the created list.

Further, when a desired file is downloaded from the host computer of the connection destination, the system further comprises a pre-read processing unit for pre-reading the file before selecting execution of the download, and among the host computers classified by the determination unit, The prefetch processing unit may download the file by giving priority to the host computer classified as having a low transfer speed based on the network distance and the reply time.

The present invention also provides a transmitting function for transmitting a predetermined measurement packet for measuring a network distance to a plurality of host computers connected to a network, and a receiving function for receiving a reply packet to the measurement packet. A distance measurement function for calculating a network distance to each of the host computers using a reply packet, a time measurement function for measuring a reply time from transmitting the measurement packet to receiving the reply packet, Another object of the present invention is to provide a storage medium storing a control program of a host computer including a determination function of classifying a plurality of host computers measured for each of the host computers using a network distance and a response time.

Here, the control program includes a RAM,
In addition to semiconductor memory such as ROM, FD, HD, PD, M
It can be stored and provided on various media such as O, MD, DVD, CD-ROM, and CD-R. In addition, the control program
It may be stored in another host computer or the like connected via a network, downloaded from the host computer, and stored in a medium such as an HD.

Further, the present invention is a network system in which a plurality of host computers are connected,
An object of the present invention is to provide a network system in which each host computer predicts a communication speed between two predetermined host computers based on a network distance and a response time measured for each of a plurality of other host computers.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail based on an embodiment shown in the drawings. Note that the present invention is not limited to this. FIG. 1 shows a configuration diagram of an embodiment of a host computer 1 of the present invention. In FIG. 1, an HTML analysis unit 101 is a part that analyzes the HTML of the URL of the designated connection destination host and extracts information of a hyperlink part therein.

The site extraction processing unit 102 extracts a site, that is, a host name part from the list of hyperlink parts received from the HTML analysis unit 101, and creates a list of sites for measuring a network distance. This list is passed to the network distance measuring unit 103.

The network distance measuring unit 103 is a unit that sends a predetermined packet to each site in the list and dynamically measures the network distance using the contents of the returned packet.

The packet transmitting / receiving unit 104 is a unit for performing processing for transmitting / receiving the predetermined packet for measuring a network distance. This packet is transmitted / received to / from a predetermined host on the network via the network connection unit 105. The reply time measuring unit 106 measures, using the timer 107, a reply time from the transmission of the packet to the return of a reply to the packet. The return time t is passed to the determination unit 108 together with the network distance n.

The determination unit 108 classifies and predicts the communication speed of each host by a predetermined calculation using the network distance and the return time. This categorized host is
It is displayed on the display unit 109. Here, for example, a host with a high communication speed is displayed prominently so as to be easily selected. Here, the network distance measurement unit 103, the packet transmission / reception unit 104, the reply time measurement unit 106, and the determination unit 1
08 constitutes the communication speed prediction unit 100. The communication speed estimating unit 100 further includes a network distance measuring unit 10.
Network distance n measured in 3 and reply time measuring unit 1
A storage unit for storing the reply time t measured at 06 may be provided.

The processing of each section of the host computer 1 is usually performed by a CPU, a RAM, a ROM, and an I / O.
The control is performed by a microcomputer including a controller, a timer, and the like, based on a control program stored in a RAM, a ROM, or the like in advance. The components of the host computer shown in FIG. 1 are necessary for realizing the functions of the present invention, and the host computer may include other components. For example, an input device such as a keyboard, an output device such as a printer, and a storage device such as a hard disk may be provided.

Hereinafter, specific processing contents such as network distance measurement via the network of the host computer of the present invention will be described. First, a method for measuring the network distance will be described. FIG. 2 shows the flow of the network distance measurement processing of the present invention. Where Cmin
Is the network distance between the two hosts closest to the host computer of the present invention. For example, C
It is assumed that min = 1.

The communication speed prediction unit substitutes Cmin for n (step C1). Hereinafter, the investigation will be performed based on the network distance n. Next, the packet transmitting / receiving unit 104
Sends a small packet that reaches only the n-th host via the network connection unit with the target host as the destination (step C2).

Here, for example, in the case of an IP packet,
If a packet of the ICMP or UDP protocol is used, it is possible to easily create a packet that can reach only the nth host. In the IP packet, a value TTL meaning that the packet is discarded when moving between the hosts by TTL is stored. Therefore, the packet may be transmitted by substituting n into this TTL.

Next, the reply time measuring unit 106 checks whether a packet has returned within a predetermined time (step C3). For example, if the packet does not return (lost) for a certain period of time (t ₀ = 10 seconds) or returns an error packet indicating that the target host cannot be found on the name server, the process ends. 2-
A].

Next, when the packet returns, the reply time measuring unit 106 checks whether the packet has reached the target host. This can be checked by checking whether data indicating that the packet has not reached the returned packet is embedded (step C).
4). If the packet has reached the target host, the return time t until the packet returns and the current network distance n are stored (step C5) [2-B].

If the target host has not been reached,
A natural number Cadd is added to n (step C6). Here, Cadd means a unit of distance when hosts are grouped by distance, and is 3, for example. The smaller the value of Cadd, the more accurate the distance measurement. Further, it is assumed that the distance is greatest when there is a host of Cmax or more between two hosts.

When Cadd is added to n, n becomes Cma
It is checked whether it is larger than x (step C7). Here, if the value is larger than Cmax, the processing is terminated because the host is too far away [2-C].

If n is equal to or smaller than Cmax, the process returns to step C2 to repeat the measurement. Here, Cmax is
This is the maximum network distance to which the transfer of the file can be completed in a desired time. For example, Cma
x is 20, but is not limited to this. Also,
The values of t ₀ and Cadd set as the above numerical values are not limited to the set values.

Next, FIG. 3 shows an explanatory diagram of an actual packet flow when the network distance n = 3. here,
The host computer that transmits the packet is Ho. If the designated host at the transfer destination is Hb, a packet is sent from the host at the transfer source Ho → Ha → Hb (paths 31 and 3).
2) A packet having the content “arrival” is returned from Hb to the host Ha → Hb → Ho (paths 33 and 3).
4). If it returns within the fixed time t _o , the process proceeds from step C4 to C5 in FIG. 2, and the value of the network distance n (=
3) and the response time t required to return.

When the transfer destination designated host is Hf, the packet is transferred in the order of the transfer source host Ho → Ha → Hb → Hd (paths 35, 36, 37). At this time, n = 3 hosts are transferred. Since the packet has passed, the packet is discarded by the host Hd. Then, from the host Hd, Hd → Hb →
From Ha to Ho, a packet "not reached" is returned (paths 38, 39, 40).

According to this measuring method, if the actual network distance to the designated host is D based on the measured network distance n, an error in which no packet is returned ([2-A] in FIG. 2) Except for the case where the packet arrives in one packet transmission (n ≦ Cmin) or the case where the packet has not finally arrived (n−Cadd <D ≦ n,
Or n> Cmax, corresponding to [2-C] in FIG. 2). That is, the approximate network distance n to the designated host can be measured.

The present invention is characterized in that the following processing is performed using the network distance n to perform classification or prediction of sites having relatively high reliability and high communication speed.

First, the network distance n is checked for all the hosts to be predicted by the method shown in FIGS. If no packet is returned or an error packet is returned, [2-A] is defined as the longest distance.

FIG. 4 shows an embodiment in which a set (n, t) of the network distance and the response time obtained in [2-B] is plotted on a graph. When the distance n ≧ Cmax, [2-C] belongs to the area (4-D) because the return time t cannot be obtained.

Here, the network distance n is N-Ca
Consider a region where dd <n <N. In this area, 4
There are three hosts -B, 4-C and 4-D. However, 4-D has a longer response time than the distance, so 4-D
-B and 4-C are considered as a different group.

That is, in order to make 4-D another group, the area of N-Cadd <n ≦ N is set to N <n ≦ N + C
The host 4 that has the fastest response speed in the add area
Divide by two in A's reply time. In the case of FIG. 4, the response time of the divided area is faster on the side (4-A) and slower on the side (4-A).
B) The area to which the host 4-A belongs is (4-c).

By classifying in this way, the host (4-B, 4-C) in (4-A) has a high response time and high reliability, and the host (4-D) in (4-A). Can be regarded as a host having a relatively slow response time but high reliability, and the host (4-A) in (4-C) can be regarded as a host having a fast response time but relatively low reliability.

The above processing is performed for the network distance of each host, and all the hosts to be measured are grouped by the network distance n and the reply time t. This makes it possible to classify or predict a host having a high transfer speed with relatively high reliability when performing a process such as file transfer.

FIG. 5 shows a flow of the grouping process. First, for all the hosts to be grouped, the communication speed predicting unit 100 measures the network distance n by the method shown in FIGS. 2 and 3 (step S1).

Next, hosts at the same network distance n are grouped into the same group (step S2). That is, the group of hosts at the same distance n is represented by G [n].
And memorize it. Next, the hosts are sorted in ascending order of the reply time t of each host for each group (S
3).

The hosts in a certain group G [n] are further divided into groups which are faster or slower than those having the quickest response time in the next farthest distance group G [n + 1]. Grouping is performed from the closest one, that is, the one with the smallest n. Therefore, G [i],
It is assumed that i = 0 and j = 1 as initial values of G [j] (step S4).

First, it is determined whether or not G [j] is a group farther than Cmax (step S5). If it is far, the current group G [i] is defined as the group B [i], and the host ([2-A] in FIG. 2) from which the error packet is returned ends as the group B [max] (step S11). .
Here, max is Cmax.

If it is closer than Cmax, G [j]
It is checked whether or not a host exists in the group (step S6). If it does not exist, add 1 to j,
Further, the next distant group is set as the investigation target (S7). If there is, the response time of the host having the earliest response time t in the group G [j] (this time is defined as F) is extracted (step S8).

Next, among the hosts in the group G [i], the hosts whose reply time is faster than the time F are group A
[N], and a slow host is set as a group B [n] (step S9). Further, j is substituted for i, 1 is added to j, the process returns to step S5, and the processes from step S5 to step S9 are repeated (step S10).

When the processing of FIG. 5 is executed for the embodiment shown in FIG. 4, the host (4-B, 4-
C) is A [N], host (4-D) at (4-a) is B [N], and host (4-A) at (4-c) is A [N
+ Cadd].

The groups A [n] and B created here
By controlling the display of the host or the prefetching of the file using [n], the reliability of the time required for the file transfer can be improved.

Next, a case in which the hosts grouped by the above processing are displayed on a WWW browser will be described. When the site extraction processing unit 102 extracts a site from the HTML and passes a list of sites other than the site currently being referred to to the network distance measurement unit 103, the groups A [n] and B [n] are subjected to the above-described grouping processing. A group is obtained.

The WWW browser displays hyperlinks of the grouped hosts in a visually easy-to-view manner, and allows a user to connect to a desired host by an easy operation. As a method of displaying the hyperlink, for example, the underline of the link in the currently referred site is a solid line, the link to the group A [n] site is a dashed line, and the link to the group B [n] site is a dotted line. As the network distance n increases, the colors become darker. Alternatively, A [n], B [n], A [n + 1], and B [n] are regarded as being closer to B [n].
The color of the underline may be darkened in the order of [n + 1].

FIG. 6 shows an embodiment of a screen display in a WWW browser. Here, L1 is a link to the site currently being referred to, L2 is a link to a site with a short distance and a fast response time (eg, A [n]), and L3 is a site with the same distance as the L2 but a slow response time ( For example, a link to B [n]), L4 is a link to a site (for example, A [n + 1]) farther than L2 and L3, and L5 is a link to a site returned with an error.

In a general browser, when no display style is specified, all underlined hyperlinks are indicated by solid lines. However, as shown in FIG. 6, the underline of L1 is a solid line, and in links L2 and L3 to sites of the same distance, the link L2 to the site with the fastest response time has a smaller underline and sites with different distances. Link L to
In L2 and L3, the link L2 to a site with a short distance is displayed with a smaller underline. L4 displays the largest underline missing. With this display, it is easy to visually understand, and it is easy for the user to select a site with the highest communication speed as possible.

Next, prefetching when downloading a file will be described. This is the WWW as shown in FIG.
In the display of the browser, before the user selects the hyperlink, the file to which the link is linked is downloaded in advance.

At this time, it is necessary to give priorities to the hosts to be read ahead due to the limitation of the number of files to be downloaded at the same time and the limitation of the storage capacity. For example, the group A is divided into groups A as described above.
After obtaining the [n] and B [n] groups, the priority of the host that prefetches the file is given the highest priority to the link in the currently referred site, and then the group B [n] and the group A [n ]
In order.

The highest priority is given to the site currently being referred to.
This is because a link in the currently referred site is likely to be continuously selected. In the case of pre-reading in the order of distant group B [n] before group A [n], the farther one takes a longer time to download, and the longer the probability of delay is, the longer the user waits. This is because the average time can be reduced. By doing so, the connection processing time and display when the user selects the site can be further reduced.

[0052]

According to the present invention, when performing processing such as reading a file between a plurality of host computers connected to a network, a predetermined packet is transmitted to the predetermined host computer, and a return time is set. With
Because we measure the distance between two hosts,
It is possible to classify and predict hosts with relatively high reliability and high communication speed.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an embodiment of a host computer of the present invention.

FIG. 2 is a flowchart of a network distance measurement according to the present invention.

FIG. 3 is an explanatory diagram of a packet flow when a network distance n is 3 in the present invention.

FIG. 4 is a graph of a network distance and a reply time according to the present invention.

FIG. 5 is a flowchart of a grouping process according to the present invention.

FIG. 6 is an embodiment of screen display in a WWW browser according to the present invention.

[Explanation of symbols]

 1 Host Computer 100 Communication Speed Prediction Unit 101 HTML Analysis Processing Unit 102 Site Extraction Processing Unit 103 Network Distance Measurement Unit 104 Packet Transmission / Reception Unit 105 Network Connection Unit 106 Reply Time Measurement Unit 107 Timer 108 Judgment Unit 109 Display Unit

 ────────────────────────────────────────────────── ─── Continued on the front page F term (reference) 5B082 FA03 HA05 HA08 5B085 AC12 BE07 BG07 5E501 AA13 AB15 AC20 AC25 BA01 BA03 BA05 DA03 EA05 EA12 FA13 FA14 FA43 5K030 LE16 MB07 MC03 5K035 BB03 CC01 DD01 FF01 GG01

Claims (6)

[Claims] A transmitting unit that transmits a predetermined measurement packet for measuring a network distance to a plurality of host computers connected to a network; a receiving unit that receives a reply packet to the measurement packet; A distance measuring unit that calculates a network distance to each of the host computers using a packet; a time measuring unit that measures a return time from when the measurement packet is transmitted to when the return packet is received; A determination unit for classifying a plurality of host computers using the network distance and the response time measured for each of the plurality of host computers. 2. The method according to claim 1, wherein the determining unit classifies the plurality of host computers according to the magnitude of the network distance, and returns a host computer belonging to the same network distance N to a minimum reply among host computers belonging to the network distance N + 1. 2. The host computer according to claim 1, wherein the host computer belonging to the network distance N is classified into two, with the response time of the host computer for which the time is measured as a threshold value. 3. An HTML analysis processing unit for extracting hyperlink information from the received data, a site extraction processing unit for creating a list of connection destination host computers based on the extracted hyperlink information, 3. The display device according to claim 2, further comprising a display unit configured to perform classification display of the host computers in accordance with a result of the classification of the host computers in the created list.
Host computer. 4. When a desired file is downloaded from a connection destination host computer, the system further comprises a prefetching processing unit that prefetches the file before selecting execution of download, and among the host computers classified by the determination unit, 4. The host computer according to claim 2, wherein the read-ahead processing unit downloads the file by giving priority to a host computer classified as having a low transfer speed according to a network distance and a reply time. 5. A transmission function for transmitting a predetermined measurement packet for measuring a network distance to a plurality of host computers connected to a network, a reception function for receiving a reply packet to the measurement packet, and a reply A distance measurement function for determining a network distance to each of the host computers using a packet, a time measurement function for measuring a response time from transmitting the measurement packet to receiving the reply packet, and each of the host computers. A storage medium storing a control program for a host computer including a determination function of classifying a plurality of host computers using the network distance and the response time measured for each host computer. 6. A network system in which a plurality of host computers according to claim 1 are connected,
A network system in which each host computer predicts a communication speed between two predetermined host computers based on a network distance and a response time measured for each of a plurality of other host computers.