Viridity Chats with IDC About the Future of the Data Center

Today, Michael Rowan, CTO and Co-founder, Viridity and Michelle Bailey, Research VP, Data Center Trends and Strategies, IDC spoke about The Future of Energy Efficiency in the Data Center.

According to IDC Research, running out of power and the rising cost of energy are primary drivers behind the heightened need to measure power consumption in the data center. Yet, surveys indicate that less than half of all enterprises measure their power usage. Why is this? It's because many IT managers don't know where to start when it comes to dynamically measuring their power usage or equipment utilization.

Michelle Bailey said that research across both existing enterprise data centers and new, emerging data centers shows that energy management remains the biggest challenge.  She presented a Top Ten list of Data Center Efficiency Strategies which are:

1. Virtualization

2. Site Rationalization

3. Operational Best Practices

4. Automation Tools and Software

5. Data Center Redesign

6. Application Rationalization

7. Modular Data Center Construction

8. Lifecycle Management and Planning

9. Rack-based Power and Cooling Retrofit

10. Architectural and Technology Refresh

Both Mike and Michelle discussed how energy resource management (ERM) software like Viridity EnergyCenter ensures greater data center efficiency whether you are extending the life of your data center or building a new one.  As Michelle said "It's hard to fix a data center problem when you can't grasp the size of the problem."  Mike demoed EnergyCenter to show how it provides a detailed understanding of power consumption and equipment utilization within hours of deploying the software.

Stay tuned to view a recorded version of this webinar which will be posted shortly on both the Viridity website and on the TechTarget site.

The Low Down on Viridity EnergyCenter Management Protocols

Viridity Software prides itself on the use of industry standard management protocols that are both managed by open standard organizations as well as vendor specific protocols. This gives us a distinct advantage because we have designed Viridity EnergyCenter to work in all data centers. EnergyCenter is vendor agnostic, meaning it will work the same with Windows servers as it will with Red Hat Enterprise Linux servers and OpenBSD servers.  EnergyCenter can even be installed on Windows and Linux.

WMI (Windows Management Instrumentation)

WMI is Microsoft Window’s implementation of WBEM (Web-Based Enterprise Management) and CIM (Common Information Model). All versions of windows since Windows 2000 have the WMI service installed by default.

The three important aspects of WMI that should be known are
Namespaces, Queries, and Security.

Namespaces (targets):

Basically you need to point at something to ask it a question. You point with a ‘moniker’. It contains the target machine and then the namespace on that machine.

“winmgmts:\\localhost\root\CIMV2”

Some namespaces are: SECURITY, directory, CIMV2, RSOP, SmsDm, and SecurityCenter. Some of these acronyms might seems familiar and that’s because they are.

Queries (questions):

To ask the namespace for something you send is a proper question such as the following.

SELECT state FROM win32_service WHERE Name=’wuauserv’

In English this is “tell me if Automatic Update is on right now.”

Security:

At a basic level you connect to another computer the same way you connect to a file share or a shared printer. If you can connect to a computer via “\\server\share” then you can read it’s WMI data. The access permissions are controlled for each namespace with the windows standard Discretionary Access Control Lists (DACLs).

SNMP (Simple Network Management Protocol)

SNMP is a standard managed by the IETF (Internet Engineering Task Force). It is a default part of any serious piece of networking equipment. It is installed on Microsoft Windows Servers but not enabled by default.

There are 3 important aspects of SNMP, Namespaces, and Security.

MIB (target)

The layout of data that you can get with SNMP is called a MIB (Management Information Base). You need to connect to an machine’s SNMP agent and then it will use a MIB to look up data on that machine.  Inside MIBs are objects organized in a tree.

Queries (questions)

To get data from a MIB you have to query for specific information in the following forms:

ASN.1:   “.1.3.6.1.2.1.25.3.3.1.2”

or

“.iso.org.dod.internet.mgmt.mib-2.host.hrDevice.hrProcessorTable.hrProcessorEntry.hrProcessorLoad”

In English, this is “Get the CPU utilization for each processor”

Security

For the most supported version, Version 2 is used. In Version 2, a Community String is used to access the SNMP service of a machine. The default Community String on most devices is ‘public’ and it is allowed read-only access. Each Community String can have it’s own access level.

Access Level	Description
read-only	The machine can only be read via SNMP
read-write	The machine can be configured via SNMP

-Joseph Paul Cohen, Viridity Engineering Team

What If IT Had To Pay The Power Bill?

Recently, BAO conducted their quarterly survey of one hundred IT professionals for us.

When they asked who has responsibility for power management in the data center they learned that it was a mix of IT, Facilities, IT & Facilities, and third party or co-location hosted facilities. This is because for most organizations the responsibility for the power bill usually sits with the Facilities department.

When asked "How are you measuring your power consumption?" about half of the IT managers asked just didn't know. 

 

 

These results are in line with information BAO has gathered from past surveys for Viridity Software. It's not unusual to learn that data centers are spending more on their energy bill than they are on capital equipment.  One reason for this is that the department ordering equipment for the data center is not the same group that pays the power bill.

What if IT paid the power bill?

If the IT department received the power bill instead of the Facilities department, they would understand the total impact of adding new equipment to the data center.  Technology investment decisions might be made based on being more energy efficient as it would impact data center ROI along with meeting business objectives.

Viridity EnergyCenter software helps both IT and Facilities managers get the information they need to monitor the power consumption and utilization of all of the networked equipment in the data center.  With this insight and actionable information, they can drive down costs and extend the life of their existing data centers. Our customers often discover 20% to 40% improvements in data center energy efficiency right away.  

So, which ever department ends up with the power bill, using Viridity EnergyCenter and putting best practices into place, they'll see their power bill get smaller instead of larger. And, that's good for the whole organization.

Discover What's In Your Data Center In < 2 Minutes

The first step towards optimizing your data center to be more energy efficient is to discover what is actually in it.  With Viridity EnergyCenter, you can view a data center inventory of assets including servers, power distribution units (PDUs), branch circuit monitors (BCMs), and more. Then, filter system types by data center, rack row, or IP address range. This video demonstrates how you can begin data center management in less than two minutes.

Once you've discovered all of the assets in this data center, you can then discover the data center power draw, events triggering jobs, the average server utilization, and the number of underutilized servers within hours.  Viridity Energy Center provides you with actionable information that results in major cost reductions while increasing your overall data center energy efficiency.

To learn more, join us on one of our upcoming live webinars where Viridity co-founder and CTO, Michael Rowan, will drill down on data center energy optimization issues, demo Viridity EnergyCenter, and answer any questions you have.

Join Us on The Road to Data Center Energy Efficiency

Lack of available power along with the energy expenses associated with IT equipment and cooling infrastructure has heightened the requirement for data center energy efficiency. Until now, it’s been difficult to successfully improve energy usage and gain visibility into hardware power usage because of the cost and complexity involved in gathering actionable information. Recently, Viridity Software’s Co-Founder and CTO, Michael Rowan explained how to get started on the road to data center energy efficiency. He discussed how to eliminate non-productive devices; optimize power and cooling; and how to optimize the data center to tie IT to the business.  Here's the presentation.

The Road to Data Center Energy Efficiency

View more presentations from Viridity Software.

Information Week Interviews Viridity Exec at Interop 2010

In April, several energy resource managment experts from Viridity Sofware headed out to Las Vegas to Interop 2010.  Michael Tresh, Viridity Software's Director of Product Management, was interviewed by Information Week at the show.  Watch as Mike discusses the efficiency gains customers are realizing when they deploy Viridity EnergyCenter in their data centers.

Getting the Right Information
for Data Center Energy Management

The industry has historically based energy planning on fundamentally flawed power calculations – typically flawed vendor faceplate power specifications, the de-rating of these specifications, or no measurement at all.  This leads to grossly inaccurate, often inflated view of the energy needed and results in margins that create unknown idle capacity.

How does this happen?  Most data center managers are in react mode most of the time as they respond to business pressures by squeezing more equipment into their facilities.  With only rudimentary tools available, they have been unable to get the information they need for power planning.

When it comes to getting the most out of the data center, the closer data center managers can get to actual, factual information, the better.  And, when it comes to energy optimization of the data center, the closest thing to nirvana is a de-rating system that bases its power data on actual, real-life system requirements and that continually improves its power requirement models through monitoring and constant learning.

Download our white paper “Intelligent, Dynamic De-rating:  The Only Way to Energy Management” to learn more.

Mike Tresh, Director of Product Management

Wikipedia Back Up After Server Meltdown

“Wikipedia back up after server meltdown” – outage Wednesday, March 23 at 2:45 pm (ET) at Wikipedia’s European Data Center….

At Viridity Software, we talk to data center managers, IT directors, facilities’ management, and industry consultants all the time.  We often share our favorite “war stories”.  Like  the one about the huge financial services company that bought over $25M (yes, million) of high-end storage and took delivery only to find out that they didn’t have enough electricity to power (or cool) the equipment.  Or most recently, an industry consultant shared his story with us about a client who experienced a data center outage across the entire facility after tripping a power distribution unit (PDU) that resulted in a cascading event simply due to not having enough available power for multiple PDUs to be in charge mode at the same time.   The list of war stories is long, but somehow you always think it cannot happen to you.

Well, apparently Wikipedia found out on March 23rd that it could happen to you.  I am sure that Wikipedia is using state-of-the-art everything in their data center to deliver their world-class website and service. But, the fact that they did have an “overheating” situation that led to a server shut down and then to failed failover (yes, a failover that didn’t work) is both unfortunate and mind-boggling.

Lets face it…the current state of IT is to react to issues versus proactively manage which results in situations like this one.  To IT’s defense, vendors are chronically pushing solutions on them that may or may not pay off, are complex to deploy, and are typically very costly.

Further, it is amazing that basic IT tools to help better understand power consumption and cooling are generally not available.  As a result, most IT pros will admit that they have no idea about consumption, hot spots, or available capacity in a given rack. 

Don’t let what happened to Wikipedia happen to you.If you are being asked (or mandated) to understand your power consumption down to the server or device, take a look at Viridity EnergyCenter software.  It installs in minutes and within an hour or two after launching your initial discovery, you will be amazed at the information you get back that can help your data center be more efficient right away.

-- Steve Keilen, VP of Marketing, Viridity Software

Data Center Déjà Vu All Over Again

Recently, I found myself shivering under a blast of cold air in the second aisle of a Silicon Valley data center.   I did not remember it being so cold when we visited it the month before. Two tiles to the left of where I was standing, the temperature was a tropical 100 degrees!  It turned out the aggressive server consolidation project was under way and had successfully eliminated two racks of old servers. Unfortunately, nobody told the facility manager that the air vents freezing the open area of the aisle were no longer cooling any servers.

Fast forward three weeks and it was déjà vu all over again. I was in a large data center in New England and beside boxes of new quad-core servers were new racks that were being added to the aisles and the last row of the data center was Arctic cold. Not surprisingly, this data center was also running a frenetic “rip-and-replace” server consolidation project  at 10 servers/week, and the facilities folks had not changed the vent layout yet.  Fast forward another four weeks to a data center in midtown Manhattan, and you guessed it, it was déjà vu all over again! New racks were being added to the end of two aisles and no changes had been made to the perforated tile configuration!

I cannot blame the data center managers. They are putting in place new, virtualized servers with greater compute power density. Unfortunately, they don’t know exactly how much power each rack and each server within the rack actually consumes.  So, how can they be expected to know how much cooling they need or what to tell the facility manager?   When a data center manager finds a convenient empty rack space to add new servers or he or removes racks of servers to reduce total power, he is being efficient as far as he can tell.

If the data center manager had calculated the power usage effectiveness (PUE), before my first two data center visits, he would have realized that even if the total server power decreased, the PUE may not have decreased to the same degree if the cooling efficiency increased with the more skewed power distribution.  Unfortunately, the data center manager did not get the return on investment (ROI) on the new servers.  

So, what’s the poor data center manager and the poor facilities manager to do? I believe the answer is “know their data center.”   They need to monitor the actual power consumed by the rack and by all equipment in it. We know that the data center manager will be making frequent adds, moves, and changes as business dictates. But without real-time granular power monitoring, the data center manager has no idea how much power they need and how to distribute it until, as one manager put it, “the fuse blows.”

Aloke Guha, Viridity Software

P.S. Last week, I talked to the facilities manager at the Silicon Valley data center. When he proudly mentioned how the data center was being progressive by reducing energy and deploying new computer room air conditioners (CRACs), I told him about the really cold zones. He got defensive and mentioned he did not how what kind of kW that IT was using. “What about real-time monitoring?” I asked.  “Exactly!” he said. “That’s what we need if we can get it.”

Data Center Metrics - "It's The Economy, Stupid"

Any article you read about data centers talks about some shade of "green," which, for the most part, is really just speak for "economy." Much like in economics, data center metrics are the language that describes the overall life cycle, health, as well as the day-to-day operations across all tiers - Facilities, IT, and Applications. Without metrics there is simply no way to communicate the essence of what a data center is, how it behaves, or how it is expected to behave.

Much like an economy, the complexities and interdependencies of the constantly moving and ever-shifting parts of the data center require persistent monitoring and tweaking just to keep it up and running. There are plenty of metrics out there to describe the economy of the data center, both macro and micro - all of them orthogonal views. There are many must reads. Two examples of the more well known ones are: Four Metrics Define Data Center “Greenness” from the Uptime Institute and PROXY PROPOSALS FOR MEASURING DATA CENTER PRODUCTIVITY from Green Grid.

Understanding all of the mechanics and motivations for these data center metrics can be a daunting task.

The metrics that describe an economy are probably some of the most successful in terms of familiarity. Why is that? Retail sales reports, unemployment figures, and GDP report on durable goods, inflation, etc.  One thing they all have in common is that they are presented relative to the previous measurement. Up is good, down is bad, or vice-versa depending on context. Simple. Looking at the actual value only means something to one with expertise in that area.

Systems that monitor, collect, and store the data that may be used to calculate metrics in the data center must help to present things in this light as well. Not only do they need to provide the data used in these constantly evolving orthogonal views that make up the data center economy, but they must also present these views to those who are not domain experts in terms of "up is good, down is bad." If we are to succeed in understanding the economy, we must make the economy understandable.

Michael Anthony