Symantec Finds Server Containing 44 Million Stolen Gaming Credentials

A Symantec blog post reports that the company recently stumbled upon a server hosting the stolen credentials for 44 million game accounts. It goes on to explain how the owners of the server made use of a botnet to process that mountain of data: "Now it's time to turn those gaming credentials into hard cash. But how do you find out which credentials are valid and thus worth some money? Three options come to mind: 1) Log on to gaming websites 44 million times! 2) Write a program to log in to the websites and check for you (this would take months). 3) Write a program that checks the login details and then distribute the program to multiple computers. Option one naturally seems next to impossible. Option two is also not very feasible, since websites typically block IP addresses after multiple failed login attempts. By taking advantage of the distributed processing that the third option offers, you can complete the task more quickly and help mitigate the multiple-login failure problems by spreading the task over more IP addresses. This is what Trojan.Loginck's creators have done."

Comments

They should post the usernames...

[BobMcD]

They could, as a service to the online community, go ahead and post the usernames that are compromised.

Re:They should post the usernames...

[JWSmythe]

    I used to have a lot of fun with that, when I was the sysadmin for a large site. It seemed every script kiddie wanted the password to it. It showed up regularly on passwordz sites. We had a whole bunch of triggers to detect and resecure accounts. One of the easy and obvious ones was to let them post it, and catch it afterwards (usually within seconds of being posted). The legitimate account holder got a notification that we changed their password to a secure one. Everyone else just sat there and wondered how we'd catch them so fast.

    That trigger was pretty low on the list though. My favorite was to catch 'em scanning for passwords. If they tried say 1000 wrong passwords in a short period, but got one or two right, we'd let them keep scannning for a while, and then block their access to the server. (iptables drop rule). Then the program would figure out which passwords they actually got right, change those, and notify the account holder of their new password. :) It was always fun to see what the delay was between them finding a password, and when it started being used from passwordz sites. In those cases, we always had the account secured before they had time to post it. The typical time from being scanned to being posted was about 12 hours. The typical time for us to reissue the passwords was less than 5 minutes.

    I can't imagine online game places wouldn't have something similar. Brute force attacks are just too easy, and people will always try them. How many different usernames can a person really try before you know that they're just brute force attacking.

Re:And if I did this...

[TubeSteak]

OK, so Symantec "recently stumbled upon a server hosting...".

What, was it placed on their doorstep one night, and they didn't notice it when they went outside to get the morning paper?

So, they wrote a crawler that intrusively scanned servers that they didn't have permission to access, opening and analyzing files that they didn't have permission to read, then published what they found?

Symantec and many other companies set up honeypot computers.
The honeypot gets infected, Symantec pulls apart the trojan and studies its web traffic.
This usually leads to the dumpsite where the trojan is uploading the data.

Many botnet/trojan masters don't bother to encrypt their data dumps or secure the server hosting it.
And even if they did, are they going to sue Symantec for unauthorized access?

Re:And if I did this...

[Monkeedude1212]

You know "IMHO" can sometimes be interpretted as "honest" and not "humble" right?