While other vendors push new storage solutions to the market, Fujitsu- Siemens Computers (FSC) thinks it has developed a more intelligent approach to the data centre - one that could actually require less hardware. Its concept, the Dynamic Data Centre (DDC), sees components working better together to deliver more efficient backup.
"DDC is our vision of how to run a data centre," says Dr Helmut Beck, vice president, storage, FSC.
"Unfortunately it is sometimes far away from reality. If you look at the data centres today you see a lot of heterogeneity, and complexity caused by this heterogeneity."
FSC aims to change that. "We support mainframes, Unix,Windows and proprietary systems. You can really attach all of your devices to a CentricStor appliance and get rid of that heterogenous environment."
The vendor has also made efforts to ensure its environment integrates with tape backups. "We are the only ones to have that heterogeneity and have that heterogenous tape storage consolidation," says Beck.
"With CentricStor we make tape fit for ILM (Information Lifecycle Management). The way to do that is to have attributes attached to groups of logical volumes. ILM means you implement differing service levels and map them on your technology."
One problem in the data centre, according to Beck, is that storage is divided into discrete areas, rather than being considered as a whole, leading to lots of independent silo architectures and an inefficient management system.
"Typically this application has very fixed, assigned static storage resources," he says. "Even if you have a big EMC or Hitachi box, you still have fixed capacity for storage.
It's very difficult for companies who need more storage. One of the consequences is because the budgets of IT departments are not growing much, they spend more and more of their budgets just operating and maintaining that environment."
This is something FSC wants to change. "Our vision is completely different to that," says Beck. "You have a certain amount of storage and resources. You can dynamically assign these resources to the services - the applications - according to the services you want and the SLAs you have with customers. You typically need a couple of technologies.”
He believes virtualisation is key to a variety of service levels and using resources efficiently. "This typically gives you a logical view of what you have. You have a number of real hard drives, but if you look at that through virtualisation you have a greater number of logical drives. By mapping such things you get the impression you have more resources. You do more with less, but you don't see the real hardware anymore.
“Many customers do their backup not in a disaster tolerant way but just in one place," he says. Again, FSC has a possible solution. "You can split one CentricStor into two sides, so you have a completely disaster tolerant solution.
“Even if one side completely goes down you still have access to the other side and can completely restore your data. The next layer is automation. You need to assign according to service levels. You also need complete integration. Not every application is able to deal with such an environment, but all the big ISVs know this is the technology of the future and have made changes."
Beck believes that the process of backing up data needs to change.
"Many of the backups in a data centre don't work. They often fail. Customers spend a lot of money on buying more tape drives and libraries.
"The time windows for backup are shrinking. People want 24/7 availability; they don't have the time to stop operations for backup. You want faster access to backup throughput. You want to stop buying all these tape drives you hardly use. You want better utilisation of the media. You want a reduction in physical devices, better utilisation of floor space, and improved availability."
John Warnants, technology and services director, Egenera, agrees that IT managers need to make wiser hardware investments when it comes to the data centre. "If you look at the average technology usage, the financial director would think that was an appalling use of IT. It's like if every time the building gets 15% full, you buy another building."
FSC is hoping to deliver better virtualisation through its BladeFrame system powered by Egenera. While FSC considered developing its own product, that route would take too long to reach the standard Egenera's products were achieving.
"They are the only company that is not only talking dynamic IT, they are delivering it," says Beck. "It's a little bit more than an OEM agreement, it's a partnership." The partnership is worth around US$300 million and the two companies will conduct joint R&D to further develop the DDC concept.
Standard Chartered Bank is one company to have implemented BladeFrame and seen results. The bank moved its highest criticality application, its global retail banking system, to BladeFrame, having previously run it on a mainframe platform and redeveloped it using IBM WebSphere and DB2.
"The system stands alone among X86 platforms in its ability to separate the identity of a server from its capacity," says Jan Verplancke, CIO, group head of technology and operations, Standard Chartered Bank.
"Before Egenera we were unable to source hardware at the high end of the stack that supported multiple environments. Instead we were locked into proprietary platforms to meet the requirements of our mission critical apps. The system delivers a combination of resiliency, cost-effectiveness and flexibility ideally suited to our vision," he adds.
Now the bank's storage system meets its needs and caters for any eventuality.Verplancke believes it offers the lowest cost, least risk choice, with the most flexibility, and means there is one less headache when planning business growth.
"Now when we're deciding whether to open a bank in a given country, technology is not on the critical path," he says.
Subscribe to Arabian Business' newsletter to receive the latest breaking news and business stories in Dubai,the UAE and the GCC straight to your inbox.