It depends on a few things. Remember, the main point of UT is to find users on the network. It doesn't make sense to run UT acquisition when users are not connected to the network.
- Are you using Dynamic UT? If so, you can greatly reduce the number of UT major acquisitions.
- Do you have lots of people coming in and out of the office throughout the day? If so, you will want to either consider dynamic UT, or run more frequent acquisitions.
- Is the network infrastructure changing frequently? If you have a lot of infrastructure changes like devices being added, trunks changing state, etc. you can limit the number of major acquistions and run minor acquisitions instead.
By default, LMS runs major acquisitions at 10 am and 2 pm. This is designed to get most users when they are in the office. If business hours dictate other times, then adjust the schedule accordingly.
Dynamic UT is a great way to supplement acquisitions and provide you near real time user updates on the network. As soon as a user connects or leaves the network, UT can be informed. Favoring dynamic UT and reducing acquisitions to one or two will greatly reduce the load on the network while giving you a chance to find "transient" users on the network.
One advantage of doing major acquisitions (over dynamic UT) is that ARP entries are resolved to hostnames during acquisition. If you go with dynamic UT, you will still want to run at least one major acquisition in the busy part of the day to make sure you get your users' hostnames showing up.
For the questions of CPU and network load, are you seeing spikes during your acquisition periods? Are these spikes causing user issues? If the answer is yes, then you should definitely back off the schedule. Either explore dynamic UT or just reduce acquisition to once or twice per day.
