This our second post in the "Agent versus Agentless" series. If you missed the first one, you'll definitely want to read it, as it provides the single biggest myth propagated by those agentless vendors:
Agent versus Agentless Myth Busting: Agented software usually places additional load on the network
Today, however, we want to cover the next biggest piece of misinformation that vendors attempt to employ during negotiations. You generally hear these things from sales guys who don't realize they are talking to technical folks who know better. If it's a technical guy from a vendor saying this stuff – you know, the one who's going to help implement and support their solution – well, maybe he's not so technical after all.
So, the second biggest myth is that an agent causes interference with the operating system and applications. This is hogwash. Here's why:
A low level agent running in the background listening to operating system events has less of an effect on a machine than executing a remote query. Posting data back to the server using HTTP is very cheap. Low resource consumption is claimed for agentless environments – which is not strictly true as the server is using technology (WMI, SNMP) on the machine to execute similar queries which causes resource utilization. An advantage of having an agent in this case is that queries can be targeted to the native operating system API and hence can result in less overall resources utilization.
Click for more information on IT Efficiency using NightWatchman Enterprise