JP6307026B2 - Web performance estimation apparatus, web performance estimation system, and web performance estimation method - Google Patents

Description

  The present invention relates to a web performance estimation apparatus, a web performance estimation system, and a web performance estimation method. In particular, the present invention relates to a technique for estimating how the performance of a website created by a website creator (the time required to display a web page by a web browser) will be in different network (NW) environments.

  Web site performance is an important indicator for Web site providers. Deterioration of Web site performance is a major cause of a decrease in the number of users. Conventionally, website performance has been measured by actual measurement.

  On the other hand, the operation of the Web browser related to the performance of the Web site is complicated, and various factors are related to the performance. For this reason, it is difficult to estimate the performance of the website (see Non-Patent Document 1).

Xiao Sophia Wang, Aruna Balasubramanian, Arvind Krishnamurthy, David Wetherall, "Demystifying Page Load Performance with WProf", 10th USENIX Symposium on Networked Systems Design and Implementation, NSDI, 2013

  However, in practice, websites are browsed in various network environments. In various network environments (for example, different wired / wireless environments, different throughput environments, and different RTT (Round Trip Time) environments), it is difficult to actually measure the performance of the website. In addition, as described above, it is difficult to estimate the performance of a website.

  An object of the present invention is to estimate the operation of a Web browser when displaying a Web page of a Web site, and to estimate the Web page display time when the network environment changes based on the estimation result of the Web browser operation. And

A web performance estimation apparatus according to an aspect of the present invention is provided:
A web performance estimation device for estimating a web page display time required when a web browser displays a web page,
In the first network environment, the request start time, which is the time when the web browser sends a request to the web server, the download processing duration, which is the time required to download the object from the web server in response to the request, A web browser operation estimation unit that obtains a size and estimates a cluster that is an object group that the web browser processes at a time based on the request start time;
A parameter related to the second network environment is acquired, the parameter is reflected in the download processing duration in the first network environment, and the first network is based on a cluster estimation result in the Web browser operation estimation unit. A web performance estimation unit for estimating a web page display time in the second network environment from a web page display time in the environment;
It is characterized by having.

In addition, a web performance estimation system according to an aspect of the present invention includes:
A Web client device that operates a Web browser;
A web performance estimation device for estimating a web page display time required when a web browser displays a web page in the web client device;
A web performance estimation system comprising:
The Web client device
In the first network environment, the request start time, which is the time when the web browser sends a request to the web server, the download processing duration, which is the time required to download the object from the web server in response to the request, Having a recording section for recording the size and
The web performance estimation device includes:
A request start time in the first network environment, a download processing duration, and an object size are acquired from the Web client device, and an object group that a Web browser processes at a time based on the request start time. A web browser operation estimation unit for estimating a cluster;
A parameter related to the second network environment is acquired, the parameter is reflected in the download processing duration in the first network environment, and the first network is based on a cluster estimation result in the Web browser operation estimation unit. A web performance estimation unit for estimating a web page display time in the second network environment from a web page display time in the environment;
It is characterized by having.

In addition, a web performance estimation method according to an aspect of the present invention includes:
A web performance estimation method in a web performance estimation apparatus for estimating a web page display time required when a web browser displays a web page,
In the first network environment, the request start time, which is the time when the web browser sends a request to the web server, the download processing duration, which is the time required to download the object from the web server in response to the request, Obtaining a size and, based on the request start time, estimating a cluster that is an object group that a web browser processes at a time;
A parameter relating to the second network environment is acquired, the parameter is reflected in the download processing duration in the first network environment, and the Web page display time in the first network environment is based on the estimation result of the cluster A step of estimating a Web page display time in the second network environment;
It is characterized by having.

In addition, a web performance estimation method according to an aspect of the present invention includes:
A Web client device that operates a Web browser;
A web performance estimation device for estimating a web page display time required when a web browser displays a web page in the web client device;
A web performance estimation method in a web performance estimation system comprising:
In the first network environment, the Web client device has a request start time that is a time when the Web browser transmits a request to the Web server, and a download process that is a time required for downloading the object from the Web server in response to the request. Recording the duration and the size of the object;
The Web performance estimation device acquires a request start time, a download processing duration, and an object size in the first network environment from the Web client device. Based on the request start time, a Web browser Estimating a cluster which is a group of objects to be processed at one time;
The web performance estimation device acquires a parameter related to a second network environment, reflects the parameter in the download processing duration in the first network environment, and based on the estimation result of the cluster, Estimating a web page display time in the second network environment from a web page display time in the network environment;
It is characterized by having.

  According to the present invention, it is possible to estimate the Web page display time when the network environment changes.

1 is an overall configuration diagram of a web performance estimation system including a web performance estimation apparatus according to an embodiment of the present invention. The figure which shows the example of the request start time and download process continuation time which the Web client which concerns on the Example of this invention records. Schematic diagram for estimating the operation of a Web browser in the Web browser operation estimation unit of the Web performance estimation apparatus according to the embodiment of the present invention. Schematic diagram when the NW environment is reflected in the operation of the Web browser in the Web performance estimation unit of the Web performance estimation apparatus according to the embodiment of the present invention. Schematic for estimating Web performance in a Web performance estimation unit of a Web performance estimation apparatus according to an embodiment of the present invention

  Embodiments of the present invention will be described below with reference to the drawings.

  In the embodiment of the present invention, based on the record of the request start time, download processing duration, and object size generated by the Web browser when displaying the Web page of the Web site, the browser An example of estimating the operation and estimating the Web page display time when the network environment changes based on the estimation result will be described. Here, the request start time is the time when the Web browser transmits a request to the Web server, and the download processing duration is the time from the request start time to the download completion time. The request start time, the download processing duration, and the object size are recorded for each of a plurality of objects constituting the web page. The web page display time is the time required for displaying a web page by the web browser.

  For example, when a Web server that holds a Web page exists in Japan, the NW environment differs between when accessing from a Web client in Japan and when accessing from a Web client in the United States. Therefore, it is assumed that the Web page display time is also different. In such a case, in the embodiment of the present invention, from the request start time when the Web client exists in Japan, the download processing duration, the object size, and the Web page display time, the Web when the Web client exists in the US Estimate page display time.

  FIG. 1 is an overall configuration diagram of a web performance estimation system including a web performance estimation apparatus 30 according to an embodiment of the present invention. The web performance estimation system includes a web client 10, a web server 20, and a web performance estimation apparatus 30. In FIG. 1, the web performance estimation device 30 is illustrated as a device different from the web client 10 and the web server 20, but the web performance estimation device 30 may be included in the web client 10 and included in the web server 20. May be.

  The Web client 10 can display a Web page held in the Web server 20 by the Web browser 101. When a URL (Uniform Resource Locator) to be estimated of a Web page is input to the Web browser 101, the Web browser 101 transmits an HTTP (Hypertext Transfer Protocol) request to the Web server 20 that holds the Web page, and the Web server 20 receives an HTTP response. In addition, the Web browser 101 sends an HTTP request to a Web server that holds objects such as HTML (Hypertext Markup Language) text, scripts (Javascript (registered trademark)), images, and the like that constitute the Web page of the URL to be estimated. And an HTTP response as a response is received. The Web browser 101 does not necessarily transmit an HTTP request to the same Web server 20, but may transmit an HTTP request to a different Web server. As described above, by repeating the HTTP request and the HTTP response, the Web browser 101 downloads an object such as an HTML text, a script, or an image from the Web server 20 and displays the Web page.

  The Web client 10 records the status of HTTP request transmission and HTTP response reception as the request start time, download processing duration, and object size. Recording of the request start time, download processing duration, and object size is performed by a recording unit (not shown) of the Web client 10. For example, the recording of the request start time, the download processing duration, and the object size may be realized by the function of the Web browser 101 itself, or may be realized by adding a function to the Web browser 101 with an extension or the like.

  FIG. 2 is a diagram illustrating an example of the request start time and the download processing duration recorded by the Web client 10 according to the embodiment of the present invention. For example, as shown in FIG. 2A, the Web client 10 displays the request start time (10: 00: 20.100) and download completion time (10: 00: 20.600) to the Web server corresponding to the URL to be estimated. Record. In addition, a request start time (10: 00: 20.200) and a download completion time (10: 00: 20.400) to another Web server constituting the Web page are recorded. The time from the request start time to the download completion time is the download processing continuation time.

  FIG. 2B shows an example in which the request start time and the download processing duration are visualized. The time when the Web page of the URL to be estimated is displayed on the Web browser 101 is the page display time. For example, some objects (images and the like) may be displayed or may not be displayed.

  The web performance estimation device 30 is a device that estimates the web page display time when the NW environment changes. The web performance estimation device 30 may be a computer configured by a CPU (Central Processing Unit) such as a processor, a memory device such as a RAM (Random Access Memory) and a ROM (Read Only Memory), a storage device such as a hard disk, and the like. For example, each function and process of the Web performance estimation device 30 described below is realized by the CPU executing data and programs stored in the storage device or the memory device.

  The web performance estimation device 30 includes a web browser operation estimation unit 301 and a web performance estimation unit 303.

  The Web browser operation estimation unit 301 acquires the request start time, the download processing duration, and the object size recorded by the Web client 10, and how the Web browser 101 downloads the object of the Web page to be estimated. And is output as “estimation result of Web browser operation”. Specifically, the web browser operation estimation unit 301 estimates a cluster, which is an object group that the web browser 101 processes at a time, based on the request start time acquired from the web client 10.

  The Web performance estimation unit 303 defines “the estimation result of the Web browser operation” by the Web browser operation estimation unit 301 and how the download processing duration recorded by the Web client 10 changes depending on the NW environment. The Web performance, which is the display time of the Web page, is estimated from the “parameter regarding NW environment”, and the result is output. Since the “parameter concerning the NW environment” differs depending on the NW environment between the Web client and the Web server, it is defined for each Web server and held in the Web performance estimation apparatus 30. The Web performance estimation unit 303 acquires the “parameter regarding the NW environment” of the Web server 20, reflects the parameter in the download processing duration acquired from the Web client 10, and based on the “estimation result of Web browser operation”, The “Web performance estimation result in a specific NW environment” which is the Web page display time is output.

  The operations of the Web browser operation estimation unit 301 and the Web performance estimation unit 303 will be described in more detail below.

  FIG. 3 is a schematic diagram for estimating the operation of the Web browser in the Web browser operation estimation unit 301 of the Web performance estimation apparatus 30 according to the embodiment of the present invention.

  The web browser operation estimation unit 301 performs clustering (hierarchical cluster rig or the like) at the request start time, and estimates a cluster that is an object group that the web browser 101 processes at a time. This estimation result is “a web browser operation estimation result”. Specifically, the Web browser operation estimation unit 301 estimates an object group whose request start time is within a predetermined range as a cluster.

  FIG. 4 is a schematic diagram when the NW environment is reflected in the operation of the Web browser in the Web performance estimation unit 303 of the Web performance estimation apparatus 30 according to the embodiment of the present invention.

  The web performance estimation unit 303 acquires “parameters related to the NW environment” and reflects the “parameters related to the NW environment” in the download processing duration of each object. “Parameters related to NW environment” includes increase / decrease in RTT (Round Trip Time), change in NW throughput, and original NW throughput when the NW environment changes. The RTT corresponds to a time obtained by subtracting the processing time at the Web server from the time from when the Web browser 101 transmits the HTTP request to the reception of the HTTP response in FIG. 2A, and the NW throughput is the HTTP response. It is inversely proportional to the time from reception until the object download completion time. Since the increase / decrease in the RTT and the increase / decrease in the NW throughput depend on the NW environment from the Web client 10 to the Web server that downloads the object, each Web server is managed by a table or the like. For example, when a Web server uses a network in which a RTT (Round Trip Time) is 5 ms and a NW throughput is 1 Mbps lower (original throughput is 3 Mbps) and a 100 kbyte object is downloaded, the Web performance is estimated. Add 5 + 100k * 8 * 1M / {3M * (3M-1M)} * 1000 (ms) to the processing duration. For example, when estimating a Web performance when a Web server uses a network in which a RTT (Round Trip Time) is 4 ms and a NW throughput is 500 kbps lower (original throughput is 1 Mbps) and a 300 kbyte object is downloaded, each object is downloaded. 4 + 300k * 8 * 500k / {1M * (1M-500k)} * 1000 (ms) is added to the processing duration.

  FIG. 5 is a schematic diagram for estimating the web performance in the web performance estimation unit 303 of the web performance estimation apparatus 30 according to the embodiment of the present invention.

  As described with reference to FIG. 4, the Web performance estimation unit 303 reflects the “parameter regarding the NW environment” in the download processing duration, and then calculates the difference in the average value of the download processing duration before and after the parameter reflection for each cluster. Calculate and estimate the Web page display time by adding the difference between the average values of the download processing duration before and after parameter reflection calculated for the cluster with the earlier request start time to the request start time of the object of a certain cluster To do.

  For example, as illustrated in FIG. 5, the web performance estimation unit 303 calculates the difference between the average values of the download processing continuation times before and after the reflection of the parameters related to the NW environment for the cluster object with the earliest request start time. For example, when 10 ms is added to the download processing duration of each object, the difference between the average values of the download processing duration before and after the reflection of the parameters related to the NW environment is 10 ms. However, since the NW environment differs depending on the Web server, the difference in the download processing continuation time before and after the parameter reflection may differ depending on the object. In consideration of such a difference in NW environment depending on the Web server, here, a difference in average value of download processing continuation time before and after reflecting the parameters related to the NW environment is calculated for each cluster.

  The Web performance estimation unit 303 adds the difference (10 ms) of the previous average value to the request start time of the cluster object having the second earliest download time. Accordingly, the request start time of the object of the cluster having the second earliest download time is the request start time acquired from the Web browser + 10 ms.

  Further, the Web performance estimation unit 303 calculates the difference (10 ms) in the average value of the download processing continuation time before and after reflecting the parameters related to the NW environment of the object of the cluster with the second earliest download time, and calculates it for the second and previous clusters. A total value (10 ms + 10 ms = 20 ms) of the difference between the average values is calculated.

  The Web performance estimation unit 303 adds the sum (20 ms) of the difference between the previous average values to the request start time of the cluster object having the third earliest download time. Therefore, the request start time of the object of the cluster having the third earliest download time is the request start time acquired from the Web browser + 20 ms.

  Similarly, the Web performance estimation unit 303 calculates a difference (10 ms) in the average value of the download processing continuation time before and after the parameter reflection related to the NW environment of the object of the cluster having the third earliest download time. A total value (10 ms + 10 ms + 10 ms = 30 ms) of the difference between the calculated average values is calculated. Thereafter, the same calculation is performed until the last cluster.

  Then, the web performance estimation unit 303 calculates the difference between the download completion times before and after the above calculation for the object with the latest download completion time that falls within the page display time before reflecting the parameters related to the NW environment, and displays the calculation result on the page. Add to time. This value is the “Web performance estimation result in a specific NW environment”.

  In FIG. 5, the difference in the average value of the download processing duration before and after the reflection of the parameters obtained for each cluster is 10 ms, but the “parameters related to the NW environment” differ depending on the NW environment where the Web server exists. Therefore, the difference in the average value of the download processing continuation time before and after the parameter reflection that is actually obtained for each cluster may differ depending on the cluster.

  As described above, in the embodiment of the present invention, the operation of the Web browser that is not affected or hardly influenced by the NW environment is estimated, and the Web when the NW environment changes is estimated based on the estimation result of the operation of the Web browser. Estimate page display time.

<Effect of the embodiment of the present invention>
As described above, according to the embodiment of the present invention, it is possible to estimate the Web page display time when the network environment changes. Specifically, based on the record of the request start time, download processing duration, and object size generated by the Web browser when displaying the Web page of the Web site, the browser operation when displaying the Web page is changed. By estimating, it is possible to accurately estimate the performance in consideration of the operation of the Web browser.

  Thereby, the website creator can discover problems related to the performance of the website provided by himself / herself by checking the performance of the website provided by himself / herself in various network environments, and correct the problems. Thus, the website provided by the user can be displayed at high speed in any network environment.

  For convenience of explanation, the Web performance estimation apparatus according to the embodiment of the present invention is described using a functional block diagram. However, the Web performance estimation apparatus according to the embodiment of the present invention may be hardware, software, or their It may be realized in combination. For example, the embodiment of the present invention causes a computer to execute a function of the Web performance estimation apparatus according to the embodiment of the present invention, and causes the computer to execute each procedure of the method according to the embodiment of the present invention. It may be realized by a program or the like. In addition, the functional units may be used in combination as necessary. In addition, the method according to the embodiment of the present invention may be performed in an order different from the order shown in the embodiment.

  As described above, the method for estimating the Web page display time when the network environment changes has been described. However, the present invention is not limited to the above-described embodiments, and various modifications and changes can be made within the scope of the claims. Application is possible.

DESCRIPTION OF SYMBOLS 10 Web client 101 Web browser 20 Web server 30 Web performance estimation apparatus 301 Web browser operation | movement estimation part 303 Web performance estimation part

Claims (7)

A web performance estimation device for estimating a web page display time required when a web browser displays a web page,
In the first network environment, the request start time, which is the time when the web browser sends a request to the web server, the download processing duration, which is the time required to download the object from the web server in response to the request, A web browser operation estimation unit that obtains a size and estimates a cluster that is an object group that the web browser processes at a time based on the request start time;
A parameter related to the second network environment is acquired, the parameter is reflected in the download processing duration in the first network environment, and the first network is based on a cluster estimation result in the Web browser operation estimation unit. A web performance estimation unit for estimating a web page display time in the second network environment from a web page display time in the environment;
A web performance estimation apparatus having:   The web performance estimation apparatus according to claim 1, wherein the web browser operation estimation unit estimates an object having a request start time within a predetermined range as a cluster.   The web performance estimation unit reflects a parameter related to the second network environment in the download processing duration in the first network environment, and then calculates an average value of the download processing duration before and after the parameter reflection for each cluster. The second network environment is calculated by adding the difference between the average values of the download processing continuation times before and after the parameter reflection calculated for the cluster with the earlier request start time to the request start time of the object of a certain cluster. The web performance estimation apparatus according to claim 1, wherein the web page display time is estimated. The parameters related to the second network environment include an increase / decrease in RTT and an increase / decrease in network throughput relative to the first network environment,
The web performance estimation device according to any one of claims 1 to 3, wherein the web performance estimation unit reflects an increase / decrease in RTT and an increase / decrease in network throughput in the download processing duration. A Web client device that operates a Web browser;
A web performance estimation device for estimating a web page display time required when a web browser displays a web page in the web client device;
A web performance estimation system comprising:
The Web client device
In the first network environment, the request start time, which is the time when the web browser sends a request to the web server, the download processing duration, which is the time required to download the object from the web server in response to the request, Having a recording section for recording the size and
The web performance estimation device includes:
A request start time in the first network environment, a download processing duration, and an object size are acquired from the Web client device, and an object group that a Web browser processes at a time based on the request start time. A web browser operation estimation unit for estimating a cluster;
A parameter related to the second network environment is acquired, the parameter is reflected in the download processing duration in the first network environment, and the first network is based on a cluster estimation result in the Web browser operation estimation unit. A web performance estimation unit for estimating a web page display time in the second network environment from a web page display time in the environment;
A web performance estimation system. A web performance estimation method in a web performance estimation apparatus for estimating a web page display time required when a web browser displays a web page,
In the first network environment, the request start time, which is the time when the web browser sends a request to the web server, the download processing duration, which is the time required to download the object from the web server in response to the request, Obtaining a size and, based on the request start time, estimating a cluster that is an object group that a web browser processes at a time;
A parameter relating to the second network environment is acquired, the parameter is reflected in the download processing duration in the first network environment, and the Web page display time in the first network environment is based on the estimation result of the cluster A step of estimating a Web page display time in the second network environment;
A web performance estimation method comprising: A Web client device that operates a Web browser;
A web performance estimation device for estimating a web page display time required when a web browser displays a web page in the web client device;
A web performance estimation method in a web performance estimation system comprising:
In the first network environment, the Web client device has a request start time that is a time when the Web browser transmits a request to the Web server, and a download process that is a time required for downloading the object from the Web server in response to the request. Recording the duration and the size of the object;
The Web performance estimation device acquires a request start time, a download processing duration, and an object size in the first network environment from the Web client device. Based on the request start time, a Web browser Estimating a cluster which is a group of objects to be processed at one time;
The web performance estimation device acquires a parameter related to a second network environment, reflects the parameter in the download processing duration in the first network environment, and based on the estimation result of the cluster, Estimating a web page display time in the second network environment from a web page display time in the network environment;
A web performance estimation method comprising:

Images