I’ve been looking at a number of different reports that would be useful to network administrators. One that came up in talking with some people about data centers is a Power and Heating Report. A number of data centers have reached power and heat loading limits, even though there is physical space left. The use of Power over Ethernet has exacerbated the problem because switches need to supply power to phones and other devices. (Yes, a switch may be in a wiring closet, which I think of as an extension of the data center and often doesn’t have the proper cooling infrastructure.)
It seems that a pretty good estimate of power and heat loading could be made by simply knowing which devices are in the facility. A network audit can identify all devices in the network and by knowing which subnets are within the facility, it should be relatively easy to determine which devices are located there. In NetMRI, I would do this by creating a device group using the set of CIDR blocks that match the subnets in the facility. Even if this isn’t exactly accurate, a good estimate is much better than guessing or having to do a periodic manual audit.
Correlate the estimate with the ambient temperature measured by the temperature probes in a set of network devices and you have a pretty good way to watch for changes and limits in populating your data center. It would be really cool to collect data from a number of data centers and be able to forecast trends in new data centers based on the collected data.
If you do periodic audits of your power and heat load, let me know. Examples of useful reports, or reports that aren’t useful if you can explain how they’re not useful, would be appreciated. We’ll take your feedback and incorporate it into NetMRI.
Re-posted with Permission
NetCraftsmen would like to acknowledge Infoblox for their permission to re-post this article which originally appeared in the Applied Infrastructure blog under http://www.infoblox.com/en/communities/blogs.html