ast 0037384 rethinking data center

7/28/2019 AST 0037384 Rethinking Data Center

1/4

When the servers in Queensborough Community Colleges data center began to show their considerableage, the college decided it was time to take a comprehensive look at how the entire data center operated,says Bryan Farr, IT senior associate at the Bayside Queens, N.Y., educational institution.

When Y2K was coming down we went to a client-server architecture, so every application had its ownphysical server, says Farr, who manages the hardware in the QCC data center.

That was in the late 1990s or early 2000. In the past ew years, as equipment wore out and we startedto have to replace a lot o servers individually, we realized that we had to optimize. We needed somethingbetter or management, cost savings and power all the main drivers.

In order to meet those needs, QCC has implemented an optimization strategy that includes servervirtualization and consolidation, storage virtualization, and substantial upgrades to the power protection

and management systems in the data center.

Key elements o the revamped IT inrastructure are an IBM BladeCenter chassis with 14 IBM HS21 bladeservers, VMware vSphere Enterprise virtualization sotware, an HP LetHand SAN and an APC Symmetrapower management system.

The Benefts o Going GreenReducing energy usage and saving taxpayer money were other important reasons or the optimization,says George Sherman, QCCs CIO.

The green aspect was important. Mayor [Michael] Bloomberg wants all city institutions to reducetheir carbon ootprints, Sherman says. But in the process o cutting power consumption, were alsoimproving our computing capabilities. Were consolidating server resources and making them more quicklyavailable, as well as creating automatic redundancy.

QCC, a campus o the City University o New York, has about 15,000 students and 1,400 aculty and stamembers. In addition to maintaining network resources or traditional classrooms as well as business andadministrative oces, the IT and Academic Computing Center sta supports the wide range o online andhybrid courses oered by QCC.

The rapidly changing needs o the academic computing environment provided an extra push towardoptimizing the IT inrastructure, says Farr.

The academic world right now changes a lot with initiatives based on grant-unding cycles, he explains.Theres always a new project coming up.

Farr and Raj Vaswani, a senior IT associate or the QCC Academic Computing Center, had been exploringthe possibilities o virtualization since 2006. But the rst long stride toward optimization came the

ollowing year when QCC deployed an IBM BladeCenter as a platorm or server consolidation. As with allthe hardware purchased or the data center, primary selection criteria were vendor support and durability,says Farr.

We asked every manuacturer that came in, What is the lie expectancy o that product line? he says.You have to make sure that something you buy is going to be around and sustainable or a while. Onemain reason we chose the IBM BladeCenter was that the model we bought was second generation, but itts rst-generation blades and now third generation. That continuity was important to us.

Synergy: Blades and VirtualizationSwitching to blade servers saved space and power, but using the blades in conjunction with VMware isthe heart o the optimization, says Vaswani. Virtualization allows ve or six virtual servers to run on asingle physical blade, and vSphere Enterprise provides a complete management platorm or the virtualenvironment.

Rethinking the Data CenterAn infrastructure optimization project saves time, money and energy at

Queensborough Community College.

T E C H N O L O G Y S O L U T I O N S
http://www.cdwg.com/


2/4

As great as the blades are by themselves, the real power comes in running them with VMware, saysVaswani. VMware with the BladeCenter is the driving orce that made the optimization successul.Without VMware the blades are very good servers, but theyre just servers.

On the other hand, using the BladeCenter really maximizes VMware eatures like VMotion [whichenables the live migration o virtual machines], Vaswani continues. Using the two technologies togethermeant that we could just wipe out all our dying servers and move them onto the BladeCenter.

QCC already owned an HP LetHand SAN, which was well suited to meet increased high-availabilitystorage needs or the virtualized environment, Farr says. The college initially bought the iSCSI-basedSAN to provide storage and backup or an e-mail server. But the SAN perormed so well that the IT starealized it could do much more, and do it at substantial savings over a Fibre Channel SAN, says Farr.

We also didnt have to pay or Fibre Channel switches and a dedicated bre network or storage, hesays. That was a big cost savings, and we were able to put that money toward actual storage.

QCC is now using 17 terabytes out o the SANs 30TB storage capacity. The HP LetHand SAN oersstorage clustering capabilities that make it easier to increase capacity. Thin provisioning eatures allocatespace as data is stored rather than requiring that storage be allocated in advance.

Backup and disaster recovery are simplied by eatures that allow administrators to access snapshots o

virtual machines or a single point in time, as well as to roll back entire volumes o backup les. The HPLetHand SAN has also decreased maintenance hassles and increased ease o use as compared to a bre-based storage area network previously used at QCC, says Farr.

With iSCSI, we take care o just about everything ourselves, he says. With the Fibre Channel system,every time we had a problem, we had to call support and wait. It was so complicated we couldntmaintain it without a lot o help.

Powering UpInitially, QCC ran two stand-alone UPS units to power the IBM BladeCenter, which contains 14 IBM HS21blade servers, eight o which are used to support a total o 60 virtual servers. One o the UPS units wasconnected to the electrical grid normally servicing the building, while the other was hooked up to abackup generator circuit.

In late 2009, the college implemented APC Symmetra power protection, a modular system that allows oreasy scaling as data center needs grow. The power management system protects the blades, along withother standard servers, the SAN and all the hardware appliances in the data center.

Our power here had been a nightmare or a couple o years. Weve had transormers blow and majorpower outages time and time again, including everything going down on campus, says Vaswani. It wasscary beore we got the APC system.

The Symmetra system also acts as a power lter and transormer, ensuring that the data center only getsclean power, which protects and extends the lie o the hardware, says Farr.

Implementation o the technology involved in the data center optimization project took place in steps overa 2.5-year period and went relatively smoothly. Backing up the virtual server environment was, however, amajor challenge, Farr says.

There was no problem backing up data rom the virtual machines, but the QCC sta was initially havingtrouble recovering snapshots o individual VMs at specic points in time rom the SAN. The diculty wasresolved with the help o ESX Ranger, backup sotware designed to be used with VMware.

Minor integration issues with some applications in the virtual environment (most notably with MicrosotSharePoint) as they were moved to virtual servers were resolved by tweaking VMware settings. Thetransition was, however, remarkably easy due in large part to VMware Converter, which duplicates physicalserver congurations on virtual servers, Vaswani says.

Nineteen out o 20 applications, i not more, were perect right away using the VMware on the blades,he says. And the converter was huge. We didnt have to rebuild servers; we just converted them tovirtual machines. It was especially great or old servers with congurations we hadnt touched in so longwe could hardly remember them.


3/4

ROI: Productivity, Time and ReliabilityQCC has invested a total o approximately $300,000 in the technology and support services necessaryor the QCC data center optimization (covering the BladeCenter with eight blade servers, VMware withVirtual Center, HP LetHand SAN and APC Symmetra).

Hard numbers or return on investment are dicult to calculate, in part because so much o theinrastructure was at the end o its lie, making it dicult to use baseline gures or comparisons, Farr

says.

The main return on investment we see is in management time, Farr says. We can get things done a lotaster. Our workload has increased because o it, but that means were doing more or more people. Wecan be much more responsive to administrators. Because were a college we dont really do ROI metrics,but we know how much better our data center works.

Increased reliability and a decreased chance o catastrophic ailures are also part o the return on theretooled data center, says Tony Hong, an IT senior associate at the Oce o Inormation Technology.

With the new system, theres less worry that systems wont come up. I there is a problem, we canusually solve it in a couple o hours, he says. In the old data center, when something didnt come up,we started praying.

QCC now does all o its application testing and troubleshooting on virtual machines at substantial timesavings, says Farr. He jokes that the VMs are so useul that he looks orward to dealing with virtual serversprawl.

That will mean that were actually using our hardware to something close to capacity, he says.

Maximum CapacityOperating close to capacity conguring the data center to make the maximum ecient use o space aswell as computing and power resources is key to data center optimization, along with having a strategyor planned growth, says Dave Cappuccio, an analyst or Gartner.

The trick with VMware, or all virtualization, or example, is not just to run it or its convenience but tolook at optimal perormance levels o the hardware and run it as close to that as you can, he says. Theaverage x86 server in the United States runs at between 10 and 12 percent [capacity], but the optimal

level may be more like 60 to 65 percent. So virtualize and put a ew images, or 30 images i that works,on a physical server.

Using SANs, virtualizing storage and deduplicating the data going into the repository are all ways toincrease eciency and decrease wasted space and energy in the data center, explains Cappuccio.

New data centers should be designed or vertical scalability, which ocuses on rack space rather thanfoor space, and emphasizes building the IT inrastructure or maximum density and eciency, Cappucciosays. Its more dicult, but the same principles can be applied to optimizing an existing data center.

First I ask, How ecient is the space you have? Cappuccio continues. One calculation I did recentlywith a client was, with increased rack density and virtualization, he could increase the computing powerin his data center by 350 percent. And he would get that increase with just a 29 percent boost in powerconsumption. That represents the optimization o all resources, including energy, which is increasingly a

big deal.

By looking at all major aspects o the IT inrastructure, QCC is headed in the right direction to maximizeperormance, comments Cappuccio.

Inrastructure optimization is usually people taking a look at things and deciding to do what theyshould have done all along, he says. You have to ask, Is there a better way to run things? It involvesundamental inrastructure design, laying things out and running them logically and eciently.


4/4

Powering the Upgrade

The benets o IT inrastructure optimization might have been enough alone to pull QueensboroughCommunity College into the process o upgrading its data center, but QCC (like many other organizations)actually had ew alternatives.

We needed to nd alternative solutions i we wanted to do more things. We couldnt expand our datacenter, and we had maxed out our power resources, says Tony Hong, an IT senior associate at QCCsOce o Inormation Technology. We had to nd a way to consolidate the servers so we would have lesspower usage.

Power and cooling are at the top o the list o issues addressed in any inrastructure optimization, saysDave Cappuccio, an analyst or Gartner. So when QCC installed an APC Symmetra power managementsystem in its data center last September, it was doing more than saeguarding operations continuity; thecollege was making sure it could continue to upgrade its inrastructure.

Organizations consolidate their servers, oten with virtualization on blades, because theyre running outo power in their current congurations, Cappuccio says. The trouble is, once you reach higher densitiesyou can ry an egg on the top o one o your racks; and you need more power or cooling. Thats whypeople bring in APC or one o the other power management systems to get that under control.

Power management systems can provide granular protection and control or various pieces and groupingso hardware, which ultimately makes optimization possible and saves energy, as well as ensuring that thedata center is up and running.

Weve nally come to the realization that we have to pay more attention to the energy we consume inIT, says Cappuccio. We can do things a lot better. We can save 15, 20, even 30 percent o energy costs

just by reconguring data centers. Power management systems protect that optimized inrastructure andmake it possible.

CALL 800.808.4239 to tALk to A speCiAList todAy.

030310

ast 0037384 rethinking data center

Documents