Every cloud has a silver lining, they say. But if a deep recession arrives in the near future, the cloud -- in the sense of remotely delivered IT from a cloud service provider -- could also be the silver lining.
In other words, what I’ll argue here is that the coming economic difficulties could be the trigger for the migration of IT services away from in-house IT departments to the cloud service provider because the reduced costs of doing so will be compelling.
So, here’s the economic backdrop. All across the Europe, austerity measures are in place amid varying degrees of political turmoil -- in Greece, Italy, Spain, Portugal and Ireland -- while the Eurozone as an entity threatens to break apart.
Meanwhile, the UK has its own austerity measures, and further afield the US is bobbing along in a depressed state with high levels of joblessness and economic malaise in public and private sectors.
With economic activity depressed, spending is tighter for all businesses, and it follows that should they be presented with the opportunity to cut costs in IT, they will take it. The cloud, with its flexible pay-for-use model, is the obvious place to do that.
But, cloud service provider-based IT has so far been unreliable. Witness Amazon's lightning-induced outage at its Dublin data centre in August and Google's e-mail outage in February. These suppliers do not offer service-level agreements and business needs certainty of IT supply before it will move any significant IT processes into the cloud.
However, now companies such as Joyent offer cloud computing with business-class service-level agreements while Nasuni does the same for cloud storage. I see this as a cloud game changer.
Contrast a company-owned data centre with racks of IT gear that has to be acquired, managed, powered, cooled and maintained, with a usage-based payment scheme for roughly the same delivered IT services from a cloud service provider. The end user gets the same kind of IT service, but the IT department no longer needs to own the kit to deliver that service in the data centre. The Capex and Opex of acquisition, management, power, cooling and support for IT goes away and is replaced by the Opex-only usage-based payment model.
Let's say there is a 10:1 saving on hardware and software but an increase in WAN costs, bringing that back to 6:1 or 7:1 overall; if a business wants to grow or simply survive in a recession, it needs to control its costs and removing up to 70 percent of IT department outlays is going to look highly attractive.
Do businesses like having IT departments? Well, it is a necessity to have IT services, but do they need data centres with their own kits in them? It’s not a great stretch to imagine that the business’ opinion about owning and operating its own or hosted data centres will harden under recessionary pressures.
Other departments will say IT has always overpromised and underdelivered. Extreme voices might say we don't run our own bank or generate our own electricity so there’s no need for our own data centre; IT should simply purchase cloud services for us.
This pressure on data centres is a huge opportunity for suppliers of cloud services and of hardware and software to cloud service providers. So, who are the likely winners?
Cloud service providers with 100 percent uptime guarantees and great compute performance should benefit. Ditto cloud storage providers that can offer similar SLAs.
Object-based storage providers and big vendors with object storage products should do well because cloud service providers will want the scalability advantages their products have over file system-based alternatives. So, Amplidata, Caringo, Scality, Dell, EMC, HDS, NetApp and others could do well.
Less obvious is the likely stance of system suppliers such as IBM, HP and others. Will they encourage a move to the cloud so they can sell more kit to the cloud service providers? Or will they try and plug the hole in the dyke through which their traditional IT customers are passing into the cloud by pushing the idea of private clouds and then the halfway house of hybrid clouds?
Cloud service providers can make more efficient use of IT equipment than customers, and a fearful view based on this premise would suggest that one cloud service provider could do the same as three end-user customers with one-third of the hardware. It follows then that a mass move to the cloud would cut the number of servers and storage arrays needed by two-thirds. That is a huge cut in the volume of the market, implies we have too many suppliers selling into the market and that not all will survive.
The opportunity view is that cloud-based services, being cheaper than DIY IT, will enable thousands or tens of thousands of SMBs to embrace the use of enterprise-class IT because it would become affordable. They would become more efficient, cloud service providers would need more equipment, and so the total global IT cake would become much, much larger.
If, or when, migration to the cloud begins en masse, it's likely to be a one-way street. Businesses offloading aspects of IT to the cloud won't want to come back to owning and operating their own data centres.
In that scenario IT suppliers to organisations of all sizes should embrace the cloud. To borrow the ice hockey phraseology, they should skate to where the puck is going to be, not where it is now or, worse, where it was. Nicholas Carr's “big switch” is happening before our eyes, and any coming recession is going to accelerate it. Are you pushing the switch or holding it back?
Chris Mellor is storage editor with The Register.
This was first published in December 2011