There are many methods by which a hacker may attack or take control of a site. I am focusing this discussion on attacks that come via form input. That is, anywhere you have input coming in from your web user, e.g. a registration form, user login or even a search on your site. Scripts could be sent to your server by entering < script> some malicious code < /script> in your input fields. The following are steps you can take to minimise risk of this happening. These measures will not make your site hacker-proof (no site can be if a hacker really has it in for you), but it can make it less of an easy target. Step 1: Place character limits on your inputs You do this by adding "maxlength" attribute into your text input tags

e.g. < input type="text" name="firstname" maxlength="15">

The example above restricts user to a 15 character input for that field. The "< script>" and "< /script>" tags alone will take 17 characters so smaller you limit your "maxlength" attribute to, harder it will be to include rogue codes in your inputs. Of course, you must ensure that you impose a suitable limit so that actual input from your valid users will not be excluded. Step 2: Filtering your data All data received from your site should be filtered, you can either filter your data when it comes into your server as user input, or when it goes out as results for your user's browser. Whether you should filter input or output, depends on your site and its requirements, there is a good discussion on this at http://www.cert.org ech_tips/malicious_code_mitigation.html/ . Filters can be written in any language, here is an example in Perl :

In point of fact, Gartner Group has recommended "that businesses hit by both Code Red and Nimda immediately investigate alternatives to IIS, including moving Web applications to Web server software from other vendors such as iPlanet and Apache".

http://www4.gartner.com/DisplayDocument?doc_cd=101034

But what about those of us who are already hosting their sites on Apache servers? I've seen lots of articles about how to protect, detect, cleanse and prevent worms from attacking IIS servers. While worms do not penetrate Apache security, they do cause damage.

Some of damage includes:Server logs get filled with junk - The Nimda worm alone created over 20,000 entries in a 2 day period in my log files.

The server is made very busy - This is especially true if you've got a custom 404 error page, as I do. This means that every time worm attempts a penetration, then entire 404 page is returned (in my case, that's about 40k). That adds up to a lot of wasted bandwidth.

I thought about this issue for a while after examining my logs and seeing thousands of 404 errors from attempted worm penetrations. Surely there was a way to at least reduce impact of these things? As I saw 404 error count increase, I realized that a significant portion of bandwidth that I was paying for was being thrown away.

An examination of log files produced several thousand attempts at each of following URLs. Obviously each of these is address of a possible weakness in an IIS server.