In environments in which you have roaming usersthat is, users who don't use the same computer every dayan important piece of information you might need to know is the machines that these users are logged on to. You need to know this information before you can remotely troubleshoot a problem (e.g., missing registry keys, missing environment variables) on a roaming user's computer.
By default, you can gather roaming users' machine information by filtering the Security logs on your domain controllers (DCs). However, this solution is time-consuming and potentially error prone, especially if you have two or more large Security logs. Instead, you can use one of two strategies to acquire this information. If you don't have too many computers on the network, you can simply poll each computer for the logged-on user. Alternatively, you can have users advertise the machines they're logged on to. By advertising I mean that the users can write to a central repository the name of the machine that they're logged on to. The advertising can be part of roaming users' logon script. Let's look at how you can implement both of these strategies and look at their advantages and disadvantages. . . .