Thursday, February 12, 2009

One World, Under F-Response

Matt Shannon is one of those guys that comes along and has a noticeable and definite impact on the world of Incident Response. First there was Nigilant32, then there was F-Response, with which Matt and his crew were able to provide a whole new...and much needed...capability to incident responders.

In October 2008, at the SANS Forensic Summit, Matt and Aaron Walters gave an excellent presentation on Voltage (F-Response + Volatility), showing how a responder can deploy F-Response and use functionality based on Volatility to fight malware. The presentation and the implementation were awesome!

More recently, Matt's shown how F-Response can be used with Nuix and Intella (be sure to check out the videos), clearly showing how F-Response's tool-agnostic framework can be used to really leverage the capabilities of other tools. Matt and Jamie Butler of Mandiant also recently announced that Memoryze supports F-Response.

Well, Matt's gone and done it again with the F-Response Enterprise Management Console (blip.tv video...even if you don't have F-Response EE yet, you should still watch this video)! Until now, pushing out F-Response EE to one system required some work in the command prompt, but now, Matt's made it so that it takes more effort for me to play BrickBreaker on my BlackBerry than it does for me to push out F-Response EE to one, two, or a dozen systems on a network!

So that's what I did...I fired up my Window 7 Ultimate Beta VM and basically followed Matt's video. As a caveat to this, I do have some experience working with F-Response EE via the command prompt, so it may have gone a bit more smoothly for me than for a first time user...but Matt's video is the best place to start. In the video, Matt walks through the various components of the UI and how you use each piece of information.

You can scan a workgroup or domain, IP address range, or directly connect to a system. I chose to connect directly to the system itself, in order to try out that functionality. Once I got connected to the VM (this is the VM I downloaded from BitTorrent, so it appears with the TuxDistro-PC system name), I could clearly see the system in the UI.

From there, all I needed to do is walk through the FEMC interface to install F-Response on the remote system and start the service with the name I provided through the UI. Since I wanted to see the physical memory (and had selected that through the UI) on the remote system, once F-Response was up and running and I had connected to the remote system, I could see two targets. The first target listed (:1) is the physical memory from the remote system, and the second target listed (:0) is the hard drive. At this point, notice that I haven't had to open the iSCSI Initiator console at all...all of the functionality for managing the use of F-Response EE is handled through a single interface.

Pretty cool so far...at this point, this has all just been a couple of mouse clicks. So the next step is to fire up the Disk Manager on the system from which I'm running all this...the memory from the remote system appears as Disk3, and the hard drive appears as Disk4. As the memory does not consist of a recognized file system, it's going to be accessible as \\.\PhysicalDisk3, and can be accessed via FTK Imager or Memoryze, which just reinforces how F-Response is a tool-agnostic platform. The larger partition from Disk4 got mapped to my analysis system as the I:\ drive, and from there I was able to run tools of my own, like RegRipper, which allows me to perform a modicum of triage and analysis.

These will show up in the FEMC UI, you don't NEED to go to the Disk Manager to see and access them...I simply show them here for the sake of transitioning from the CLI method. The physical disks themselves are shown in the Connect tab (you may need to refresh the UI to see them, but they'll be there). Also, you can follow what's happening in the Messages tab, and even cut-n-paste all of the messages from your session and use those in your case notes as your documentation.

Matt blogged on the FEMC here.

I've got two words for you. Suh. WEET! There's no question...for what amounts to a very modest price, you (consultant, IT staffer, etc.) have the capability to get answers NOW, whether you're responding to an incident, or addressing an HR issue, or even just network or system troubleshooting. For example, the current model of incident response is when something happens, many organizations call someone; from there, it takes time set up a conference bridge, describe to someone who isn't familiar with your infrastructure what's going on, and then they have to grab their gear and get on a plane...it could be 24 (or more) hours before someone's on-site, and then they have collect data. With F-Response and the FEMC, you can now preserve the data (i.e., in the case of new malware, collect a memory dump), provide it to consultants, and start getting answers while whomever is designated to come on-site is still looking for flights!

No comments: