Published: September 20 2008, 08:48 AM | 1 Comment(s)
by Chris Stakutis

Your data center is by far the largest energy consumer in your facility. Data centers grab a stunning 2% of the overall energy used in the U.S. today and are growing their usage the fastest (doubling within the next 5 years). Yet, most facilities do not measure  the total power going into the data center and virtually none measure the power on individual circuits and outlets. 

But wait! We’re here to tell you that this is EASY to do, and you don’t even have to disrupt power to your systems in order to do it!  If you have an ultra-modern Power Distribution Unit (PDU) you might already have network access to this information.  But even if you don’t (and we didn’t in any of our data centers), it is very straightforward to buy and install power-metering gear.  Essentially, you clip a “split core” ring onto each circuit wire at the PDU (there are a variety of different ways to do this). These hardware circuit-boards are very inexpensive and each can handle 80 or 100 or more circuits.  With this simple addition to your PDUs, we are now live monitoring 1,000 circuits in our main data center.

 And don’t stop there! Inside each rack, you probably have a long power strip (probably dual) that power all the servers and equipment, right? New versions of these strips can monitor each individual outlet.   

Once you have the monitoring gear in place, you can typically pull the data via SNMP over IP.  We at CA have recently enhanced our core monitoring software product to understand how to pull this new data and have built new applications for display and management. It is so important to first measure measure measure!!  Why make any changes if you don’t know where you are today?  And you certainly want to be able to show real data and report on performance improvements!!!

Chris J Stakutis
VP Emerging Technologies

Christopher.Stakutis@ca.com

Share this post:  Email

Share

Published: September 05 2008, 01:37 PM | 1 Comment(s)
by Sam Somashekar

Everyone these days is talking about saving the planet. Natural resource depletion is at an all-time high, and is evidenced by today’s events such as the oil “crisis.” The notion of going “Green” means something different to everyone. Does it mean putting more money in your pocket or someone else’s pocket? How about saving a tree or planting a new one? What about reduction of solid state and paper waste? Or, reduction of space used for equipment?

“Green” can actually mean all of the above. It really depends upon who you’re talking to. For an IT person, “Green” typically manifests itself as spending less for energy and cooling and optimizing the physical space they have to use for IT equipment. In speaking to a variety of people, it appears that the notion of saving on the bottom line is the initial driver of implementing Green practices. And, if this helps them become better corporate citizens then that becomes a big plus. Understanding what it means to go “Green” is an important first step to figuring out how to get there.

So, how do you get there? For IT people, this typically involves investment in new energy-efficient hardware technologies. It also means consolidation of existing physical space for servers and other IT equipment and leveraging technologies such as virtualization. However, it takes money, resources and time to do this. Budgets may not exist due to the state of the economy today. And, after you invest in that new energy-efficient server, can you be sure that you will actually be using it more than 20% of the time so that you can get value out of your investment? It’s important to understand that new hardware and services alone cannot guarantee effective implementation of Green strategies. There needs to be a way to properly utilize new or even existing hardware at their peak capacity. And, if they are not needed for a period of time to know when to shut them down to save on energy and cooling costs. Consolidation and virtualization are great strategies as well, but if you cannot effectively manage the virtual machines (VM) that are sure to grow rampant in an organization then you may be at a loss. It’s important to control VMs just as you now control physical machines in terms of configurations, usage and compliance. Going Green should not cause more headaches. And then there’s the mainframe, a great Green platform in and of itself; but again, just by running on the mainframe doesn’t guarantee that you are effectively taking advantage of its features.

This is where software comes into the fold. Software becomes that bridge between solid-state and business practices, a way to directly support the optimization of existing investments and new investments. Software becomes a mechanism by which you can measure how Green you really are at any given moment, and that’s important to ensure you are not just spinning your wheels. Going Green affects all aspects of IT and the business, and should be supported in conjunction with other business initiatives to prevent potential silo situations which over time create additional overhead and complexity. For example, to maximize the investment in new energy-efficient hardware as part of an existing pool of server resources, software can ensure that work is scheduled intelligently across that pool. Potential extensions can allow for integration with energy monitoring software that may come with new hardware, which can enable event consolidation at the server level with other IT management software. And, this is just the tip of the iceberg, as you will see in future blog entries.

So, what does “Green” mean to you? I invite you to let us know and your methods for getting there.

Share this post:  Email

Share