Using an agent for a server power management solution is the only way to identify whether useful work is being carried out on a server. This is the only way to accurately determine if a server is being used, enabling you to easily discover and decommission the 15% of servers doing no useful work.
There is also no solution for true useful work detection with an agentless approach as this data is not exposed remotely.
Missed a post in this series? If you’ve ever heard a vendor tell you that agentless is better than agented, you’ll definitely want to catch up on this series.
Part 1: Agent versus Agentless Myth Busting: Agented software usually places additional load on the network
Part 2: Agent versus Agentless Myth Busting: Interference with the operating system and applications
Part 3: Agent versus Agentless Myth Busting: Agents open up the machines to security vulnerabilities
Part 4: Agent versus Agentless Myth Busting: Agentless means there is no agent at all
Part 5: Agent versus Agentless Myth Busting: Protecting user data