I've accepted an answer, but sadly, I believe we're stuck with our original worst case scenario: CAPTCHA everyone on purchase attempts of the crap. Short explanation: caching / web farms make it impossible to track hits, and any workaround (sending a non-cached web-beacon, writing to a unified table, etc.) slows the site down worse than the bots would. There is likely some pricey hardware from Cisco or the like that can help at a high level, but it's hard to justify the cost if CAPTCHA-ing everyone is an alternative. I'll attempt a more full explanation later, as well as cleaning this up for future searchers (though others are welcome to try, as it's community wiki).

情况

这是关于woot.com上的垃圾销售。我是Woot Workshop的总统,Woot Workshop是Woot的子公司,负责设计,撰写产品描述,播客,博客文章,并主持论坛。我使用CSS/HTML,对其他技术几乎不熟悉。我与开发人员密切合作,在这里讨论了所有的答案(以及我们的许多其他想法)。

可用性是我工作的重要组成部分,而让网站变得令人兴奋和有趣则是剩下的大部分工作。这就是下面三个目标的来源。验证码损害了可用性,机器人从我们的垃圾销售中偷走了乐趣和兴奋。

机器人一秒钟就会在我们的首页上猛击数十次屏幕抓取(和/或扫描我们的RSS),以寻找随机垃圾销售。他们一看到这个,就会触发程序的第二阶段登录,点击“我要一个”,填好表格,然后买下这些垃圾。

评价

lc:在stackoverflow和其他使用此方法的站点上,他们几乎总是处理已验证(登录)的用户,因为正在尝试的任务需要这样。

在Woot上,匿名(未登录)用户可以查看我们的主页。换句话说,撞击机器人可以不经过身份验证(除了IP地址之外基本上无法跟踪)。

所以我们又回到了扫描IP, a)在这个云网络和垃圾邮件僵尸的时代是相当无用的,b)考虑到来自一个IP地址的业务数量,捕获了太多无辜的人(更不用说非静态IP isp的问题和试图跟踪它的潜在性能影响)。

还有,让别人给我们打电话是最糟糕的情况。我们能让他们给你打电话吗?

布拉德克:内德·巴切德的方法看起来很酷,但它们是专门设计来击败为网络站点构建的机器人的。我们的问题是机器人是专门用来破坏我们网站的。其中一些方法可能只在很短的时间内有效,直到脚本编写人员将他们的机器人进化为忽略蜜罐,从屏幕上抓取附近的标签名称而不是表单id,并使用支持javascript的浏览器控件。

 

lc再次说道:“当然,除非炒作是你们营销计划的一部分。”是的,绝对是。当物品出现时的惊喜,以及当你设法得到一件物品时的兴奋,可能比你实际得到的垃圾一样重要,甚至更重要。任何消除先到/先得的东西都不利于“赢”的快感。

 

novatrust:就我个人而言,欢迎我们新的机器人霸主。我们实际上提供RSSfeeds,允许第三方应用程序扫描我们的网站的产品信息,但不是在主站HTML之前。如果我的理解正确的话,你的解决方案通过完全牺牲目标1来帮助目标2(性能问题),并放弃机器人将购买大部分垃圾的事实。我给你的回答投了赞成票,因为你最后一段的悲观情绪对我来说是准确的。这里似乎没有什么灵丹妙药。

其余的响应通常依赖于IP跟踪,这似乎是无用的(僵尸网络/僵尸/云网络)和有害的(捕获许多来自相同IP目的地的无辜的人)。

还有其他方法/想法吗?我的开发人员一直在说“让我们只做验证码”,但我希望有更少的侵入性方法,让所有真正想要我们的垃圾的人。

最初的问题

假设你卖的东西很便宜,但有很高的感知价值,而你的数量非常有限。没有人确切地知道你什么时候会卖这个东西。超过一百万人经常来看你卖什么。

你最终会发现脚本和机器人试图通过编程方式[a]找出你何时出售该道具,[b]确保他们是第一批购买该道具的人。这很糟糕,有两个原因:

你的网站被非人类攻击,拖慢了所有人的速度。 编剧最终“赢得”了产品,让常客感到被骗了。

一个看似显而易见的解决方案是为用户在下单前设置一些障碍,但这至少有三个问题:

The user experience sucks for humans, as they have to decipher CAPTCHA, pick out the cat, or solve a math problem. If the perceived benefit is high enough, and the crowd large enough, some group will find their way around any tweak, leading to an arms race. (This is especially true the simpler the tweak is; hidden 'comments' form, re-arranging the form elements, mis-labeling them, hidden 'gotcha' text all will work once and then need to be changed to fight targeting this specific form.) Even if the scripters can't 'solve' your tweak it doesn't prevent them from slamming your front page, and then sounding an alarm for the scripter to fill out the order, manually. Given they get the advantage from solving [a], they will likely still win [b] since they'll be the first humans reaching the order page. Additionally, 1. still happens, causing server errors and a decreased performance for everyone.

另一种解决方案是经常监视ip攻击,阻止它们进入防火墙,或以其他方式阻止它们排序。这个可以解2。和阻止[b],但扫描ip对性能的影响是巨大的,可能会导致更多像1这样的问题。比编剧自己造成的还要严重。此外,云网络和垃圾邮件僵尸的可能性使得IP检查相当无用。

第三个想法,强迫订单表单加载一段时间(比如半秒),可能会减慢快速订单的进度,但同样,脚本编写人员仍然是第一个进入的人,在任何速度下都不会对实际用户造成损害。

目标

将道具卖给非脚本人。 保持网站运行的速度不被机器人减慢。 不要让“正常”用户完成任何任务来证明他们是人类。


当前回答

对于你的特定问题(而不是一般问题),一个潜在的解决方案是,如果用户想要看到“垃圾”,就要求他们登录。只向恰好登录的用户显示垃圾奖品。所有其他项目都可以对非登录用户保持可见,就像他们一直以来所做的那样。然后你的忠实用户就会优先考虑这些垃圾。

很明显,你必须通知你的用户这一点,也许会通知他们这是为了增加真正的用户发现垃圾的机会。

如果你的具体问题是机器人收集某种特定类型的物品,那么就采取限制性最小的替代方案,只防御这种特定的攻击。这个选项可以防止验证码和您所关心的可使用性问题。

如果机器人登录并开始发送垃圾邮件,您可以强制他们注销并锁定帐户。

如果他们只是为了得到一袋垃圾,他们会很快离开,你的页面不会受到大量的点击。忘掉那些高度技术性的解决方案吧。

其他回答

You should have some record of the users who have purchased BOC most often, why not just ban those accounts or something. Sure legit users will be banned in this process but you are a business providing a product and if your are being abused by a group of users and such you have the right to refuse service to them. You have a lot of info on your users including paypal and bank accounts, you could ban those accounts forcing the bot users to get new accounts. Certainly I could come up with a script to buy BOC all the time or just download one from the net, but I have better morals than that. Never actually having successfully purchased BOC, I know the frustration of legit users who would like to receive a BOC in the hopes of getting a great deal. Perhaps instead of offering a BOC as an individual item every once and awhile, you could just give it to random users every day. When they receive an item they get a little note and and an additional item saying they also received a BOC. Then the only way someone could get a BOC is if they legitimately purchased something that only an actual human would have wanted. There would be nothing better than purchasing a coffee maker or something and also receiving a 42" tv or something in addition to your legitimate purchase. I think the majority of script kiddies would no longer be interested in your site if in order to get a BOC they would also have to commit to a purchase of more than 10 dollars.

开发一个首页组件和购物车,它们不能在浏览器中本机运行。如果你使用Flex/Flash或Silverlight之类的东西,就很难抓取,而且你可以完全控制服务器通信,因此可以完全屏蔽脚本编写者的内容。

我有一个解决方案(可能)没有列出(因为我还没有读完这里所有的……)

您可以通过浏览器的User Agent字符串跟踪唯一用户。从本质上讲,通过检查哪些信息是可用的“唯一”,你将能够获得足够的信息来区分不同的人(即使是在相同的IP地址上)。

看看EFF写的这篇文章 以及这个网站(也由EFF),将“测试”你的独特只是基于你的用户代理从浏览器。

为了获得更好的唯一性,你可以在唯一性和ip地址的信息位之间进行比较,从而真正得到罪犯/机器人的可能性。


也从EFF签出这个pdf

两种解决方案,一种高科技,一种低科技。

First the high-tech: The BOC offerings sell out in a seconds because bots get many of them in the first few milliseconds. So instead of trying to defeat the bots, sell them what they are scanning for: a bag of crap. Worthless crap, of course: bent paper clips and defiled photos of Rosie O'Donnell. Then have built-in random delays on the server for a few seconds at a time. As the sale continues, the actual value of the product sold will increase while the sell price does not. That way the first buyers (bots in the first few milliseconds) will get something worth much less than what they paid (brown onion cakes?), the next buyers (slower bots or faster humans) will get something unspectacular but worth the purchase price (bought on consignment?), and the last buyers (almost all humans) will get something worth more than the purchase price (break out champagne?). That flat-screen TV might be in the very last BOC purchased.

任何等待太久的人都会错过机会,但与此同时,任何购买太快的人也会吃亏。诀窍在于等待一段时间……但不要太多。这里面有一些运气,这是应该的。

低技术含量的解决方案是将中国银行的名称改为人类可以理解但机器人不能理解的名称。酒皮屎?装着臭味的麻袋?与各种商品相邻的拓扑平面?不要重复使用相同的名称,使用稍微不同的图片,并在产品描述中解释实际销售的产品。

Restrict the times at which you release offers: For example: only from 7 minutes to 8 minutes past the start of an hour. Do not deviate from this, and give penalties on the order of a couple seconds to IPs which check a lot in the half hour before the release time. It then becomes advantageous for bot owners to only screen scrape for a couple minutes every hour instead of all. the. time. Also, because a normal person can check a site once every hour but not every second, you put normal people on a much more even footing with the bots.

饼干: 使用仅由唯一ID(数据库表的键)组成的跟踪cookie。对没有cookie的客户端、无效cookie、使用新IP的相同cookie的客户端或频繁使用cookie的客户端给予“发布延迟”。

识别可能的机器人: cookie将导致机器人为它们控制的每个IP请求多个cookie,这是可以跟踪的行为。只有一个发出的cookie的ip很可能是普通的客户端。有许多发出cookie的ip要么是大型NAT-ed网络,要么是一个机器人。我不知道你会如何区分这些,但公司可能更有可能拥有DNS服务器、网页之类的东西。