Sumo wrestlers are powerful. American football players are agile. But only rugby players are powerful, agile and lean enough to illustrate the strengths of hyperscale cloud datacenter infrastructure. Learn how to get lean with hyperscale and how disaggregated hardware elminates overprovisioning.

twosumoandareferee-2400px.png

Sumo wresters are very powerful

After Kisenosato Yutaka won the New Year Grand Sumo tournament, he was crowned yokuzuna, the first Japanese Sumo grand champion in 19 years. Kisenosato is 30 years old and weighs 174 kg (385 lbs). Sumo wrestlers have to be powerful.  You can see pictures of him in action here.

American football players are powerful and agile

american-football-1445331_1280.jpgIn America, Marcus Cannon plays for the New England Patriots. He weighs in at 154 kg (340 pounds) and plays offensive tackle. For those of you unfamiliar with American football, an offensive tackle has to be powerful, but also agile.  You can see photos of him here and here.

 

Rugby players - powerful, agile and lean

Ericsson-cloud-hyperscale-blog-rugby.jpgSumo wrestlers and NFL players deserve our admiration, but my favorite contact sport is rugby. To play rugby, you not only have to be powerful, you not only have to be agile, but you have to be lean. And you have to do it on both offense and defense. No sitting on the sidelines sucking on oxygen.

Alesana Tuilagi, a Samoan rugby player for the Newcastle Falcons, only weighs 125 kg (275 pounds), but as you can see from this video clip, he demonstrates all three qualities:

Oh. And he has a brother.  Check him out here.

If you enjoy scaring the heck out of your desk-bound self by imagining how you would react to one of these guys barreling down on you, check out 15 of the Most Jacked Beasts in Rugby.  And rugby is equal opportunity fierce - check out Seven INSANE tries from women's rugby in Langford.

If you'd rather pursue a more gentle activity, read how Ericsson Hyperscale Datacenter System 8000 can make you lean. Well, your infrastructure. If you, personally want to be powerful, agile, and lean, you'll have to start playing rugby.

How to get lean with hyperscale datacenter infrastructure

Traditional IT must deal with a difficult conundrum: to put your business in a position to respond to both opportunities and challenges, it has to overprovision. If, instead, you opt for a lean infrastructure to contain costs, it leaves you with a lengthy provisioning process that makes you slow to respond.

What you need is a way to assemble systems to provision workloads on demand. That capability is provided primarily through a software-defined infrastructure. A software-defined infrastructure allows you to buy and operate only the hardware components you need, dramatically reducing the need to overprovision. It works by defining, in software, the configuration of compute, storage, and network resources that is best suited to run a particular workload with a specific set of customer requirements.

Once you have defined your virtual infrastructures, you can provision your workloads by spinning up instances of that infrastructure. And when the workloads have been completed, you can “disassemble” the instances and return those resources to the common pool.

This approach leads to higher utilization rates compared to dedicated hardware that is selected to support peak usage, resulting in over-provisioning. Since you are getting more use from the hardware you have, your utilization rate goes up while your capex goes down. Because you have fewer under-utilized, shadow systems, a lean infrastructure also drops your costs for power, cooling, and real estate.

Disaggregated hardware eliminates overprovisioning

The disaggregated architecture of Ericsson Hyperscale Datacenter System 8000 virtually eliminates the need for overprovisioning on the one hand, and lengthy procurement processes on the other.

Disaggregation allows you to purchase only the individual components you need to pursue your business opportunities. Since compute resources have shorter lifecycles than storage resources, you can save money by upgrading only your compute resources. Or, since you don't have to pay for an entire server, you can upgrade to the latest compute resources sooner. This approach also makes the business justification for procurement less onerous.

As a result, you can run your business on the components best suited to the needs of your customers, whether those components are the latest innovation or less costly legacy systems. Not only do you get best-in-class return on investment, but you can amortize components such as disks, chassis, and cables that don’t need to be replaced as often as other resources.

Since you are getting more use from the hardware you have, your utilization rate goes up while your capex goes down. Fewer idle systems and less over-provisioning consumes less power, which in turn reduces your utility costs.

To explore our vision of hyperscale further, check out our white paper, which sets out the key considerations and questions for datacenter operators at every layer – from hardware to applications.

Download the white paper

For more information

About the photograph

I took the photograph of that C7 Chevrolet Corvette at a Corvette gathering in Denver, in the summer of 2016.

 


Cloud Infrastructure Digital Industrialization

Rick Ramsey

I started my high tech training as an avionics technician in the US Air Force. While studying Economics at UC Berkeley, I wrote reference manuals and developer guides for two artificial intelligence languages, ART and SYNTEL. At Sun Microsystems I wrote about hardware, software, and toolkits for developers and sysadmins, and published “All About Administering NIS+.” I served as information architect before joining BigAdmin, which morphed into the Systems Community at Oracle. I left Oracle in May of 2015, and now write for Ericsson.

Follow Rick Ramsey:
Rick Ramsey

Discussions