Sunday, November 12, 2006

Digital Forensics Documentation

One aspect of digital forensics that is often overlooked by a number of folks is documentation. If you've ever taken an class in incident response or digital forensics, undoubtedly you've heard about the need to properly document your work. Really, the thousand-foot goal with documentation is to provide an audit trail of what actions you took, (and sometimes) why you performed those actions. Ultimately, you may be required to defend the actions you took, and the documentation of your actions is likely to play a role.

A few of the different reasons why documentation of your actions is important are:
These aren't the only reasons, just some of the more common ones. The first bullet point deals with the concept that both sides should get a fair chance to examine the evidence. For instance, if your report supports a theory that a suspect executed a malicious executable on a specific system, the areas that you analyzed (e.g. prefetch directories, access times, etc.) are likely to be of interest. If you document the different aspects of the evidence that you examined, then another examiner can recreate the substantive content of your actions, and examine the same areas.

This leads to another question that I frequently get asked, and it is how much documentation is adequate? The answer is that it depends on who your audience is. There is the adage that says "write your report so that it can be used in court" (or some other similar wording). The problem is that this adage isn't really helpful. One good idea that I've come across is to document your work so that someone else, with a similar knowledge level, can recreate your steps. There are two implicit notions in this statement that are worth clarifying: the similar knowledge level, and the lack of mentioning tools.

A person with a similar knowledge level, would ideally be able to arrive at similar statements as you. The catch is, what is reasonable for a similar knowledge level? If you take a look at the syllabuses for a number of digital forensics courses, you'll notice that they cover a lot of the same ground, file systems, operating system specifics, etc. It would be fair to assume that someone who is at a similar knowledge level would have learned the "basics". On the other hand, if you have come up with some new technique, it might be prudent to explain the technique in a bit more depth (or at least reference it if the technique is published elsewhere). Typically, the more in-depth explanations can go in an appendix, although it's not a strict requirement, more a matter of style.

The lack of mentioning a specific tool is also noteworthy. It might be a good idea to document what tool you used, but someone else isn't required to use the same tool to examine the same evidence. The idea behind this is that, assuming a tool works in a correct manner, and the output from the tool was interpreted in a manner that was intended, another person should be able to arrive at similar conclusions using their own methods. This is why it's important to understand how your tools work behind the scenes. You don't necessarily need to know that your specific tool stores the various MFT entries in a linked list (although it's possible you might). Understanding what the tool does, and it's limitations however are likely more important.

Another common issue is what to put in your final report. The answer is that it really depends on who your audience is, and ask them. For instance, if your work supports corporate security, find out from them what it is that they want. They may say they only want a few specific things and that's it. If you're working for an attorney, they will likely have a specific strategy, and be able to provide you with enough guidance with regards to what it is they want. It might be they only want a list of the names of files created within a specific time. In this case, a spreadsheet of the names of the files would likely be sufficient. For matters of style, you might want to format the report look a little more professional than just a spreadsheet (e.g. letterhead, readable font, etc.) Again the choice is up to you.

Labels: ,


Wednesday, November 08, 2006

What CSI does right

I was at a training class last year and the instructor made a good point about the TV show CSI (Crime Scene Investigation). While the actual techniques/methods/etc. the show uses may not always be accurate with respect to real life (some are, some aren't), the characters do perform a lot of experiments. If you don't know the answer to a question, you can do some research and hope there is an answer somewhere, or you can perform some experimentation yourself.

Labels: ,


This page is powered by Blogger. Isn't yours?

My blog has moved! Redirecting…

You should be automatically redirected. If not, visit http://www.forensicblog.org and update your bookmarks.