Building private Clouds is the way that most enterprises will move as their IT requirements develop and grow. That's the view of Martin Snellgrove, Solutions Principal for Infrastructure Consulting at EMC.
While it may seem a self-evident truth to some, it does identify some of the underlying issues that face not only the larger, established enterprises but also the established systems vendors as they collectively contemplate moving from the increasingly costly traditions of on-premise information management to the more agile, flexible and cost-effective world of Cloud Computing.
“Most of these businesses are aware what the potential advantages of the Cloud will be, once they have got there,” Snellgrove said, “but the process of transition from where they are now is often difficult for them to understand.”
This is certainly a major problem for those enterprises – particularly those with well-entrenched business processes that work well in an on-premise environment. But it is also, as Snellgrove acknowledged, something of an issue for the systems vendors as they contemplate the change in market dynamics that the Cloud is likely to bring.
The issues for the users are the obvious ones – the costs of continually upgrading hardware and software systems can be painful enough these days, particularly when it comes to software costs and associated issues such as training and maintenance budgets. These alone make on-premise computing increasingly punitive. But the big lever now is the cost of energy and suitable real-estate and support facilities for the needed new datacentres; that is now a major management headache. These are major drivers in the move towards the Cloud, coupled with the increased flexibility and business agility that comes as part of the package.
Private Cloud has emerged as an accepted next step for the larger enterprises, partly because the phrase helps make the move seem more manageable and controllable than moving into the unknown, unbounded badlands of `the Cloud’. But even in the more controlled perception of the private Cloud, most enterprises are going to require help and guidance to make the transition, which ahs the potential to be a two-edged sword for many enterprises.
At one level is can be a real benefit, and to that end Snellgrove has developed a road map aimed at guiding established businesses through the processes of moving to a Cloud-based – or at least Cloud-like – infrastructure. This starts by setting out some of the fundamental questions that need to be asked by business and on-premise IT managers – such as how do they go about consolidating and virtualising their existing infrastructure. This may seem a very low level question to some but for many enterprises it still represents a first hurdle that is yet to be jumped.
It then sets out the key activities that need to be undertaken as businesses progress from a physical datacentre to a virtualised one, on to the addition of a private Cloud of external resources and the final stage – the end of the Cloud-beginning so to speak – the seamless mix of on-premise, external private and external public resources and services that is the Cloud. These include early steps such as rationalising the applications in use (because many will in practice be redundant), transforming the current data protection strategy and defining current IT services and costs.
Next comes the implementation of security and compliance controls, the planning of network convergence and the migration of legacy applications, followed by qualifying Cloud vendors, extending security and compliance policies and controls, extending the mobility of the workload to exploit public resources and tuning applications to optimise Cloud performance.
“This has to be done at two different levels,” he said, “one aimed at the needs of business managers and one for the IT departmental staff. The latter are often working under pressure at tactical level which can make it difficult to appreciate the issues surrounding a strategic change like a move to the Cloud.”
But while expert support is important, this can raise a question mark over the position of some of the experts. Some, after all, have a vested interest (aka business model) that is geared to being the systems provider to business users, at a time when their need to actually require or own more physical systems is coming into question. Snellgrove is aware that EMC and its new best buddy in the Cloud, Cisco, both run the risk of being bracketed in that way.
Such companies run the risk of promoting the notion of `private’ Clouds as actually being a users’ very own datacentre 'out there somewhere'. In practice such infrastructure development may utilise Cloud-like architectures, but the users will not be out in the Cloud. Such a model will most likely end up with much of the Capex and Opex investment that would be needed to expand an on-premise installation. Snellgrove acknowledges that the business model for the vendors is going to have to change, and suggested that EMC is already planning for it.
“From the vendors’ point view the Cloud is all about delivering automation and commoditization,” he said. “Servers are already being commoditized by the hypervisor in virtualised datacentres, and the next target will be the commoditization of the hypervisor itself.” He acknowledged that these admirable goals of automation and commoditization bring with them a consequent reduction in profit margins for the vendors. So in a market where margins are unremittingly tending towards zero he did admit that he does see a change in business model coming for EMC, but declined to be drawn on speculating what that might be.
However, as it is clear that service providers themselves are an obvious target market for the company, building a business model that taps into annuity revenue generation seems an obvious candidate. This is also a market where there is potential for co-dependency to build, for the service providers could well help themselves by promoting and reselling the consultancy services Snellgrove’s team at EMC is now offering.
Post new Comment