Browse by Tag

Subscribe by Email

Your email:

JouleX Blog

Current Articles | RSS Feed RSS Feed

Agentless or Agent Based - What's the difference?

 

Enterprises trying to find the right energy management solution for their businesses need to know what the agent based/agentless difference actually means to their bottom line. We know that both solutions are equally accurate for gathering PC and physical server energy data. Beyond that, the benefits and capabilities of the agentless solution outshine agent based in every area.

Agent Based energy management solutions require that software be installed on every device to measure the energy from it. For example, to measure and report the energy from a PC, an agent based solution would have to be installed on that PC. Then you can query the software and it would return the energy usage of the device.  Since the only devices you can realistically install monitoring software on are PCs and servers, NO agent based energy management solution is able to measure, monitor or manage the full scope of your enterprise’s devices.

In most companies, servers and PCs are only 50% - 70% of their IT infrastructure. You can’t load software on the peripheral equipment, such as switches, routers, wireless access points, and printers; therefore you have no ability to get energy data using a software agent.  With the JEM agentless energy management approach, these barriers don’t exist.

Agent Based Devices Managed:

JouleX Energy Manager Devices Managed:


  • PCs
  • Physical Servers
 
  • PCs
  • Macs
  • Thin clients
  • Servers
    • Physical
    • Virtual
  • VoIP phones
  • Monitors
  • Printers
  • Copiers
  • Scanners
  • Network switches and routers
  • Wireless access points
  • Heating, ventilating, air conditioning, and lighting systems
  • And more

Even if agent based solutions could be installed on every peripheral device in an enterprise, there are other significant challenges to that approach.

  • COST- It costs time and money to install and manage agents on every desktop.  If, for example, your company has 50,000 PCs in its environment, the costs grow exponentially. Consider the amount of time (X 50,000) to:
    • Download and install
    • Deploy and test
    • Monitor and measure
    • Patch and update
    • Upgrade
  • SECURITY– An agent requires software to run in a privileged mode. If there is a vulnerability or problem with the agent itself, the machine is exposed to additional security vulnerability, whereas the agentless approach uses the existing supported Operating System based metrics. This means, the system is not unnecessarily exposed to additional security vulnerabilities.
  • DEVICE SUPPORT – By leveraging that core operating system functionality, with the agentless approach, tracking agent and operating system releases and hardware releases is not necessary.
  • DISCOVERY – The JEM agentless system collects an inventory of all assets that are drawing energy on your network. Basically that means “it just works.”  There is no guess work or double checking to ensure that none of the machines are overlooked or forgotten.

One last, but key, distinction between the agentless and agent based solutions is this: Agent based software uses additional energy on every single device it’s installed on, increasing the very costs your company is trying to measure, monitor, and control. 

When you look beyond the surface it's plain that these solutions are only similar in name, but not in capabilities and effectiveness for the enerprise.

For more information, read: Questions Answered about Agentless vs Agent Based

 

Thank you for taking the time to visit the JouleX IT Blog. We hope you'll join us on Facebook or Twitter and subscribe to our RSS Feed!  We look forward to joining you in the Green IT conversation!

Comments

Would you mind to elaborate on how "agent based software uses additional energy on every single device it’s installed on"? 
 
When you don't have agent, you actually use an "embedded agent" (call it SNMP, Windows WMI, or whatever). Enabling and using that function is the same as installing an agent. 
 
Lastly, I wonder how enabling Windows Management Interface (WMI) on Windows (and leaving that port open) is perceived as more secure that an agent based solution? 
 
Thanks for reading, 
Guy 
Posted @ Wednesday, August 10, 2011 3:36 AM by Guy Kawasaki
Hi: 
It was really interesting to follow your article right until the same sentence the previous comment signals: I think it trashes all trace of seriousness you could have. Can you put in money the sum you are "wasting" by installing an agent in a PC? Do you mean that you could go to the extreme to disable your screen driver to save energy, or uninstall your printer drivers for the same? Would you remove your RAM sticks when surfing the web and then plug them when you open Excel? Please, I appreciate your effort to communicate the advantages of not having to deploy an agent but I think you are really stretching it too much (and not giving your article more value in doing so). I´m sure that not having an agent can also present some challenges and would be fantastic if you can inform the people about them (I think about persistence without the network, scripting to make things happen, security holes to open to make protocols talk to each other, etc).
Posted @ Monday, September 19, 2011 11:59 AM by Juan Pablo Garcia
Juan Pablo,  
 
Thank you for your comments. I have addressed these issues as well as Guy's questions which are very similar on another blog entry. I would love your feedback on http://www.joulex.net/Green_IT_Blog/bid/60790/Questions-Answered-about-Agentless-vs-Agent-Based 
 
Thanks for the comments!
Posted @ Wednesday, September 28, 2011 6:24 AM by Mark Davidson
Post Comment
Name
 *
Email
 *
Website (optional)
Comment
 *

Allowed tags: <a> link, <b> bold, <i> italics