Top three web server vulnerabilities

August 3, 2006 —

A few weeks back, I wrote about old worms and attacks that were still propagating around the Net. I received feedback from folks who said that only in the last six months or so have they seen the death of the Nimda and Code Red worms for the most part. I did some checking and most of my systems now only get a Code Red hit about once a week or less, while Nimda hits have all but disappeared on most of my monitored network segments.

These days, the web server top three seems to be ASN.1 exploits, scans for PHP applications and ongoing scans for the Horde application framework. These are almost continual, with ASN.1 hitting about the same rate as scans on our honeypot systems for open spam relays. I'm surprised by this fact because I have a hard time believing that there are web servers out there that still have not been patched and cleaned up from the ASN.1 debacle.

For those organizations using PHP on Internet exposed systems, I expect they feel the true brunt of the ongoing attacks. So much attacker attention is being paid to PHP that it must feel like a never-ending game of patch and counter patch. PHP has its place, and it is a powerful language, but it leaves something to be desired in terms of a history of secure development.

I also hear more security pundits saying that the day of the webworm is over. This statement shocks and amuses me by its shortsightedness. True, we have developed some mechanisms for managing the outbreaks of web-based worms to a larger extent, but the likelihood of seeing more is pretty high. The targets are just too interesting and the web server and web applications are just too plentiful to be ignored by attackers.

Will we ever see another Nimda or Code Red? Many shake their heads no, but it's possible. Application-based worms are likely, and we are only seeing the beginning in things like the PHP worms now crawling around.

So, the bottom line is: don't get complacent and forget those lessons that the earlier nasty worms taught us. In the meantime, would the owner of all those ASN.1 infected systems apply the patch please?

ITworld.com