Spammers Still Relying on Centralized Web Servers

In so many ways, spammers have mastered the art of massively paralleled, massively redundant computing.

Instead of setting up or hijacking a few email servers, they have created software that takes over thousands of computers for the purpose of sending spam against the will of their owners.

85% of the computers they're sending from might quickly be blacklisted and blocked by most internet provider's spam email blocker software. However, this doesn't really matter. They have so many computers sending so many email messages - thousands or millions of them will still pollute inboxes around the planet. A few suckers will purchase from their campaigns, either making them money directory or via an affiliate program. All of this at no cost to them.

Perhaps this is evil, but it's also brilliant.

However, it seems this approach does have an Achilles' heal. Many spam campaigns still rely on a single or handful of web servers. Once these servers are taken offline, messages that refer users to them won't result in profits for the spammer.

The server may fail under the increased load a spamming campaign can generate. Other times, the provider of connectivity may catch wind of what's going on and pull the plug.

So why don't spammer just use the same computers they're sending spam from to host their web pages? Sometimes they do, but often they can't. Most desktop systems connected to the internet are shielded from outside traffic using a technology called Network Address Translation or NAT. The computers can easily send traffic outwards (like spam), but generally can't receive and fulfill requests they didn't make (like a web server does).

Back