This page contains a Flash digital edition of a book.


Where the box fits

Fully specified deployments of standalone data centers within standard ISO shipping containers remain thin on the ground. But there are whispers that deployments are taking place beyond the early adoptions by universities and research centers around the world

In recent times FOCUS has asked about commecial deployments of containerized dat centers, but names and case studies have been hard to come by. That said, there have been clues that maybe data centers in these form factors are being deployed commercially. HP has an un-named large industrial manufacturing customer which it says has deployed a high-performance computing solution in its POD (Performance Optimized Data Center) in three phases – a project it says started in 2009 and is coming to completion now.

We also found another HP deployment in Western Australia. Research establishment iVEC says it has entered the prestigious ranks of the top 100 supercomputers in the world thanks to its HP POD installation at its Murdoch supercomputing facility.

The 107 Teraflop system (1 Teraflop = 1trn floating point operations per second) uses HP ProLiant Blade servers with 9,600 cores and 500 terabytes of high-performance storage. It will be part of iVEC’s data network, which operates at 10 gigabits per second.

SGI said one of its ICE Cube systems was chosen by deciBel research – a high tech company running the US Army’s Joint Land Attack Cruise Missile Defence Elevated Netted Sensor System. It chose two 20-foot SGI Ice Cube data centers to use for its simulation and evaluation exercises and performance testing for the program. DeciBel research president and founder Dr Bassem Mahafza

A Verari container housing part of NASA’s Nebula hybrid-cloud infrastructure

says its choice was not only about keeping equipment cool. “Other key Ice Cube factors include its highly efficient cooling system as well as its ability to comfortably accommodate a small IT team within the data center itself. It’s ten-ton chiller not only cools the onboard computer equipment but also the data center technicians who will work in the communications and operations areas within ICE Cube.”

In Spain a colocation company called Adam developed its own container called Adam Databox. The firm said it couldn’t find a suitable product on the market so developed its own using standard rack and in-row cooling components. It is deploying the containers at its own colocation facility just outside Madrid and said it is bringing them to market.

The US Government’s National Aeronautics and Space Administration deployed part of the infrastructure that makes up its hybrid cloud called Nebula inside a container made by a San Diego, California, firm Verari. The box’s total power draw is 350kW. It has about 15 equipment racks inside with an in-rack cooling set-up. The unit is connected to external chillers. There are several compute

clusters inside, made by a variety of vendors including Cisco’s Unified Computing System, a Silicon Mechanics cluster and a cluster of Verari blade servers.

All networking in Nebula is 10Gb Ethernet. The architecture uses core network switches supplied by Cisco and network cards by Intel and QLogic. The platform is using KVM virtualization technology.

In addition to providing services to NASA internally, Nebula enables its users to utilize resources deployed in a commercial vendors’ public clouds when appropriate, such as the times when data requested is public and does not have to be deployed on the private infrastructure.

Strangely, IBM hasn’t announce any new container products so far this year. Instead it announced that its Portable Modular Data Centre (PMDC) was launched in Singapore. In the announcement it said its first customer in Asia Pacific is an electronics manufacturing company in Malaysia, proving commercial deployments are taking place, and in numerous fields. The problem, however, is finding those willing to discuss how and why deployments are taking place. n


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20