Event Log Managment

Logs .. Logs and More Logs

Do you need to track who/where/when for activities done against the OU’s in your AD?

With Windows 2003 those were difficult questions to answer, we could get some very basic information from Directory Services Auditing; but it was limited and you had to read through several cryptic events (id 566).  With the advanced auditing settings with Windows 2008 R2 you can get some better information (you can do this same thing with Windows 2008 but it has to be done via command line and applied every time servers restart).

I don’t want to bore you with Windows 2003 auditing or the command line options for Windows 2008 Domains (if you need them, I will get you the information).  So let’s just jump right to using Windows 2008 R2, because we can now apply the advanced auditing settings via Group Policy.

Now when you turn on the Advanced Audit Policy Configuration you are turning OFF the basic or standard Audit Policies.  The Advanced Audit Policy Configuration allows you to control what AD will audit at a more granular level.  Now for the focus of this discussion we are only going to talk about setting up auditing for activity on our Domain Controllers, the other systems in your environment will be a different discussion.

So where do we start so that we can answer our question at the top of this discussion?

First, turn on the correct auditing.  Open up Group Policy Management Editor and drill down as seen in Fig 1.  **Take note of the green highlight.

GPO to Track OU changesFig 1

For this discussion we are focusing on DS Access and its subcategories.  We only want to turn on Audit Directory Service Changes, see Fig 2.  This category only generates events on domain controllers and is very useful for tracking changes to Active Directory objects that have object level auditing enabled. These events not only tell you what object and property was changed and by whom but also the new value of the affected properties.

GPO part 2Fig 2

Now that we have step 1 completed, setting up AD for auditing, it’s time to configure WHAT we want to audit.  This next step is done via Active Directory Users and Computers.  Open up the properties of your AD and drill down to setup the auditing for Create and Delete Organizational Unit objects as seen in Fig 3.

Fig 3

Now we need to add more granularity so we need to do this process 1 more time and this time instead of checking boxes on the Object tab we are going to check 2 boxes on the Properties tab, see Fig 4.

Fig 4

Now that our auditing is setup what type of events can we expect to see?

Here are a few examples:

In this example (Fig 5), id 5137, we see an OU being created by the Administrator.

Fig 5

Figure 6 shows a Sub OU being created.

Fig 6

Figure 7 shows id 5139, an OU being moved.

Fig 7

Now for the best one, this one comes as a pair of messages – OU rename, part of id 5136.

Figure 8 shows the first part of the rename process.

Fig 8

Figure 9 shows the second part of the rename process.

Fig 9

Now let’s contrast all of this with an event that is part of the good old standard auditing.   Let’s take moving an OU; with the Advanced Auditing we get id 5139 (fig 7), nice and easy to read and understand.  Now here is id 4662 that you would get for the same thing with standard auditing, fig 10.

Fig 10

With standard auditing some of the other items that we looked at would be next to impossible with auditing, such as tracking when an OU is renamed and as you can see from fig 10 hard to read and understand if you did get an event.

Now if your AD is in Mixed Mode (W2k8 and W2k3) you are stuck with standard auditing.

August 16, 2012 Posted by | Audit Policy, Audting, Directory Services, Event Log, Windows 2008 | , , , , , , , , , | Leave a comment

Logging, Logging

Here lately we’ve been hearing a lot about Stuxnet and Duqu.  Well this week is no different, but there is some insight into how one of these could have been slowed down if not prevented.  In an article released on Nov 30, 2011 at eWeek.com, Duqu Attackers Wiped All Linux CandC Servers to Cover Tracks, there are some good insights even though it’s almost just a side note.  The article mainly discusses what the Duqu attackers did to help cover their tracks.  The part of the article that caught my eye was the last paragraph.

Even though there was a possibility of a zero-day, the researchers thought it was more likely that the servers’ root passwords were brute-forced, based on a log of a user attempting to log in as root multiple times over an 8-minute period from an IP address in Singapore before finally succeeding.

It all comes down to logging and then doing something with those logs.  Having a solution (www.logrhythm.com) that can help point out these types of items is critical.  Just to have logging turned on gets you nothing, you need to do something with these logs.  Having a solution in place and watching for anomalies like this could have helped to slow Duqu down.  Logs won’t stop the attackers but if we are being proactive and watching (with the correct solution) we would at least slow them down some.

As this also shows that attackers aren’t worried about people logging their activities because they know most organizations/admins aren’t doing anything with the logs until well after the fact.

One of the things that you need to be asking your SIEM/Log Management provider is how would they have picked up on something like this?  Can it be correlated/detected that it was coming from the same IP address and repeated attempts?  Most SIEM/Log Management vendors will tell you “sure we can do that”.  Well my next question would be: What if this was across multiple Operating Systems, could you detect it then?  Now for my shameless plug, with LogRhythm the answer is yes.  Some vendors expect you to be a subject matter expert across all Operating Systems, network devices and applications.  With LogRhythm you don’t need to be, our normalization handles that for you.  We can take being proactive to the next level with advanced correlation and SmartRemediation.

December 2, 2011 Posted by | Audting, Event Log, Hacking, Log Management, Uncategorized | , , , , , | Leave a comment

Inside and Outside Hack Attempts

Over the last several years I have conducted quite a few webinars with Randy F. Smith on a variety of topics that relate to Windows Audit Policies and Log Management.  Two of these truly drive home the point about why you need to be looking at your logs (not just Windows but all sources; *NIX and Network Devices as well).  The first of these was conducted on Jan 20, 2009 entitled “Anatomy of a Hack: Tracking an Intruder with Security Logs” and most recently on Feb 4, 2010 entitled “Insider Gone Bad: Tracking Their Steps and Building Your Case with the Security Log “.

March 3, 2010 Posted by | Audit Policy, Audting, Event Log, Hacking, Log Management | , , , , , , , | Leave a comment

Tracking Down Audit Policy Changes

Yesterday I held a webinar about how to track down changes to your Audit Policy.  I have had several requests for the recorded session link from the people who attended.  So I thought I would share the webinar with everyone else.  To view the webinar please visit:

**Feb 14, 2011; Do to some unforseen issues at Prism I can no longer in good faith promote their product or services and I have removed all links to their website.

January 17, 2008 Posted by | Audit Policy, Audting, Event Log, Log Management | , , , | Leave a comment