An increasing “virtual” perspective on storage, where it is of less importance where the data is physically stored, has led to the outsourcing of storage and its management, particularly in a disaster recovery and business continuity context. Redeal, a distributor of electrical products that is part of the global Rexel Group, is one of the growing number of companies that have taken this approach.
Peter Jameson, Redeal chief information officer, says the company relies significantly on its IT systems for the successful running of the business and the requirement for a reliable back-up facility is paramount. But, it also did not want to deal with the cost of substantially idle disaster recovery facilities.
Working with Computer Concepts in Christchurch, Redeal implemented virtualis-ation which ensures the right amount of resource (storage or processing) is available to cope with varying needs from a large number of applications.
The company runs a duplicate SAN, one in its Auckland office and an identical configuration at Computer Concepts, with 300 to 400GB of capacity on each.
This provides back-up chiefly for Redeal’s Unix ERP system. The back-up facility has not had to be called on for a genuine emergency yet, “though we were close to it in the recent Auckland power cuts”, says Jameson. But a full online test was run in May this year, turning the whole Unix ERP workload successfully over to the Computer Concepts facility then bringing it back smoothly.
A ‘holistic view’
Enterprises should address their storage concerns in the context of a larger strategy, advises analyst firm Forrester.
Forrester surveyed companies in North America and Europe early this year and found storage now represents around 19 per cent of IT hardware budget – more than either systems management or network hardware. This growth spend is expected as organisations need to increase their storage capacity and meet a raft of compliance requirements, which could also drive up the costs of associated back-ups and data replication, says Forrester.
Forrester says a company-wide strategy that “encompasses networked storage, data retention, and data protection” will pay off in the medium to long-term. To ensure future-proofing, the strategy should be “independent of any one vendor’s biases and view of the market”.
“A formal and comprehensive storage strategy would not only give context to decisions regarding specific products and new initiatives,” reports Forrester analyst Andrew Reichman, “but it would also lay the foundation for transforming the storage organisation from a reactive department that manages technology to a proactive, service-oriented organisation.”
Saint Kentigern College has put such a storage pool into practice across its two campuses in Auckland.
With the use of ICT being central to the College’s approach, it aims to give its students 24x7 reliable computer access and quality of service in whichever area of the campus they may be. “Giving the students an extra edge is tremendously important to us, so many of the ICT initiatives that we undertake are designed to ensure that we create that advantage for our students,” says Walter Chieng, director of ICT.
The challenges of running two independent campuses with a large, complex and growing ICT infrastructure led Saint Kentigern to undertake a full review of the organisation’s infrastructure earlier this year.
The three-year ICT plan includes consolidating some ICT services between the two campuses. The key aims of the consolidation project are to reduce costs, remove duplication of services and manpower, share resources and engineering between the two sites and provide shared connectivity to resources located in the two sites.
As part of the project, the College needed to consolidate all of its server and data storage requirements. This was to provide robust, reliable systems that would offer full redundancy and ensure data protection for the College, with the reliability and scalability required for “mission-critical” applications.
St Kentigern chose its storage provider with the help of its existing ICT partner, Infinity Solutions. The emphasis was on data protection, reliability, performance and scalability?–?and cost. “When you’re dealing with an education institution, dollars mean a lot and justification of the decision you make and the money you invest is absolutely critical,” explains Chieng.
When Infinity proposed NetApp, Chieng investigated further. “I talked to my industry peers who were working in banks and other mission critical environments and found [many] had implemented NetApp solutions.
“This gave me the confidence that the solutions would work well in our demanding environment and would provide the levels of protection we require,” he says.
To achieve its aim of a simple, scalable and cost effective consolidated processing system for the two campuses, Saint Kentigern implemented blade server technology connected to a NetApp IP SAN via iSCSI to store key Windows application data, Exchange stores and SQL Server databases using the iSCSI protocol.
It has been almost a year since the product was first deployed and the College is delighted with the results. “Both Infinity and NetApp have invested a lot of time and engineering hours configuring and optimising the system, and educating us in terms of understanding what we are doing to ensure we optimised the facilities right from the start,” says Chieng.
Chieng says the NetApp IP system was deployed, configured and operational in approximately four hours. On top of this, NetApp’s flexibility in volume size allows the configured volumes to grow dynamically and use the space available, without administrative intervention.
The combination of blade centre and NetApp IP SAN has integrated seamlessly into the college’s existing infrastructure. This saves staff time and resources for the college, says Chieng.
NetApp’s RAID 4 configuration is designed to offer protection against data loss caused by disk failure. NetApp SnapShot technology enables the creation of instant file recovery via Snapshot and instant file system recovery via SnapRestore providing a combination of disk and data protection.
“The fact that the system is now automated has given me a higher level of confidence in the back-end to service Saint Kentigern’s requirements, as well as the faith to test and demonstrate the system when necessary,” says Chieng. “So when your key executives come in and say ‘how do I know there is redundancy’ you can demonstrate on the fly and know it will be successful. To be able to do that is just fantastic.
“It’s been about achieving data efficiency and data integrity. It’s about workflow and optimising the operation, using the resources smartly and not duplicating anything across the campus. We have received a very significant ROI and can now reassign resources into other locations to service our users.”
Trends and consolidation
Forrester points out a large number even of major organisations in the US are still using direct attached storage even in the face of obvious economies.
Networked storage obviates the problem of running out of storage in a particular location or running out of the kind or storage that a particular application needs while less suitable capacity lies idle. The technology also facilitates centralised management. A large amount of storage in one location lends itself to organisation in a hierarchy of cheap-slow and expensive-fast media.
Increasing automation in storage manage-ment follows a similar well-established track as in many other parts of IT, says Roger Cockayne, managing director of storage specialist Revera. In the first stage, resources have to be assigned manually?–?“people wouldn’t buy large disks because they weren’t assured of being able to use all the capacity effectively. Five to six years ago, they got over that. The allocation is now done through software; but still some people try to do it by hand.” Eventually, he says, the capability will be incorporated in hardware.
The storage allocation systems promulgated by IBM and EMC use a standalone processor, typically a Linux box, while those from Hitachi, Sun and HP put the software in the disk controller, says Cockayne.
There are encouraging signs organisations are willing to spend up front on sophisticated storage management in order to save on ongoing operational cost, but the trend is not moving as quickly as perhaps it should, he says. “There are real savings in storage, given that it can’t be swapped in and out as easily as servers can; typically it’s with you for five or six years, so it has to be made to perform.”
Some vendors, meanwhile, are busy inte-grating themselves with a view to becoming one-stop shops that will make the client more comfortable working with a single supplier. EMC, with a blizzard of acquisitions, has set itself up to cover the whole of “information lifecycle management”.
The company has made 23 acquisitions in the past 24 months, but all are logical and directed at enabling the customer to deal with just one vendor for all storage-related needs, says NZ manager Craig Stoddart.
Stoddart notes from the customers’ point of view, “Information has changed from something that was stored in the cellar with the IT team to become an integral part of the business; something the CEO wants to keep a finger on constantly.”
More CEOs are conscious of the need to be able to recover information, even information that is several years old, either for legal compliance or for the sake of the corporate memory. At the same time, information is growing rapidly and often unexpectedly. “Grenades” of recent decades include multi-Mbyte images from digital cameras, which can throw off storage growth predictions.
Networked storage obviates the problem of running out of storage in a particular location, or running out of the kind of storage that a particular application needs while less suitable capacity lies idle. The technology also facilitates centralised management. A large amount of storage in one location lends itself to organisation in a hierarchy of cheap-slow and expensive-fast media. Businesses “are recognising that you don’t need to put everything on Tier 1 storage,” says Stoddart. Less often needed data can be sidelined to lower-cost media.
Today’s automated systems, says Cockayne, can be trusted to do the task and to move the data associated with a certain application to a different kind of storage if it is not performing adequately.
• How to gain reliability, scalability and ROI from flexibly managed storage.
• Why storage providers are remaking themselves into ‘one-stop shops’
• What are new storage related services like virtualisation and security.
Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.