A little countermeasure of the impact statistic of anti-brush IP flow software in jquery _jquery

Encounter brush IP, brush flow software puzzle

This thought that the effective control on the same day on the same IP only once the scheme can effectively prevent users to refresh the address of the phenomenon of cheating, I have to admit that the inexperienced--and have been ignoring the current brush flow software so powerful, our project also encountered the use of brush flow software, resulting in a lot of garbage data, Even affect the accuracy of statistics.

In order to solve this situation also deliberately downloaded two well-known high profile brush flow software "traffic treasure" and "Flow Wizard" to understand, do not know, a good thing is really a.

These two software principles are the same, it is estimated that the other is the same, is the use of the principle of network exchange, using the network node geographical differences, the user randomness, so that the effect of the real effective access, that is, your computer hangs a brush flow of software, your address will be also in the hang of all users visit, Of course, people help you brush at the same time you also help others brush, all this kung fu software in the back to help you complete, hanging a few minutes immediately can see the flow of slowly come up, said a lot of nonsense, the following also said countermeasures.

Response Plan

In view of some netizens do not like to see the full text on the next break theory of the situation, here to emphasize that all the background data submitted have to do IP check , and this article is discussed in this case is based on the Brush IP response scheme.

Scenario One: Asynchronously submitting data via Ajax (invalid)

At first the promotion of the address is clicked in the background parsing page is to record the IP, time and other information, this way is obviously difficult to prevent brush flow software, so consider the form of asynchronous submission of data through Ajax.

In the beginning or underestimate these rogue software, think just through a simple simulation HTTP request does not trigger JS script, so the first scheme is completed through the page after loading, through Ajax asynchronous submission of record requests, the results are invalid, the experiment proved that this way only for the relatively low-level robot effective;

Scenario Two: Determine the requested client browser window width or height (not valid)

Through the scheme one can speculate that these flow software is not a simple simulation of HTTP requests, that is, through the Real browser request, but I hang the machine "was" to help others brush the flow at the same time, did not see any web page open, only through the grab Bag tool to see non-stop requests, Then I guess the brush flow software is not hiding a browser window, or set the browser window is very small ... To this guess I through JS to determine whether the client is currently open browser window area is greater than a certain value (such as width 300px, high 200px, I believe that no one will use such a small area to see the site Bar ~), more than a lower value to submit data through Ajax.

The result is still invalid, also deliberately wrote a small program to record the browser type of each request, the size of the window ..., the result makes me have no language, the request of the browser viewable area are very normal, and even a lot of resolution than my monitor is still high, self despise ing ...

Scenario Three: Use mouse events as a basis for normal access ( effective )

After several experiments, the conclusion is that these robots are not simple, but after all, the robot, then consider the use of mouse events to determine whether the robot, such as MouseMove, MouseDown, MouseOver, Of course, by letting users choose to click the button and so on to consider the operation as a basis (of course, to consider the operation experience is good), put a simple script below:

Copy Code code as follows:

<script src= "Jquery-1.4.1.min.js" type= "Text/javascript" ></script>
<script language= "javascript" type= "Text/javascript" >
$ (document). Ready (function () {
var movetimes = 0; Number of moves
$ (document). MouseMove (function (even) {
movetimes++;
if (Movetimes > 100) {//event is executed more than N, set a more conservative value
$ (document). Unbind ("MouseMove"); Unbind MouseMove
Asynchronous submit data is executed here
Alert ("Submit record request asynchronously!");
}
});
});
</script>

Summarize:

After many attempts, finally in the third scheme to achieve results, but also do not rule out the continuous improvement of these robots in the future to break through this simple test link, after all, while outsmart. But robots, after all, is a robot, it may be possible to increase the number of artificial operating conditions to judge, it is estimated to prevent such brush IP flow software impact statistics. Perhaps due to the specificity of the project, these programs are not necessarily suitable for all situations, after all, these brush flow software are mostly grass-roots webmaster use and are usually brush their own website:), here also hope that the netizen in the yard also put forward good suggestions, how to deal with these brush flow software, share the experience.