Hedging Against the Cloud Trap
Like so many technology megatrends, the promise of the cloud continues to exceed the real-life experiences of IT organizations whose companies were built when on-premises data centers were the only real option.
The latest eye-opening analysis of the true costs of large-scale cloud reliance caught our attention in an article on the Blocks & Files site headlined “Public cloud Hotel California costing exposed.”
Veteran tech journalist Chris Mellor began with a startling revelation: “Storing a petabyte of data in an Amazon file system can cost 400 percent more per month than storing it in an on-premises all-flash file store, and moving it back on premises can cost tens of thousands of dollars. Cloud suppliers want to stop customers moving their data out.” Mellor was citing research conducted by William Bain financial analyst Jason Ader, who told his subscribers: “As a general rule, based on our research, any amount of production data (data that needs a minimum level of performance) above one petabyte will be substantially less expensive to host in an on-prem data center versus the cloud.”
Over and above the monthly charges, prohibitive egress costs form the kind of lock-in that customers thought they were avoiding when they first took the plunge. The hyperscalers have created a pricing structure designed like a lobster trap, to lure customers in smoothly while making it very expensive to repatriate workloads back to the data center or shift them to another cloud. Like the cold-water crustaceans we love to enjoy every summer, those terabytes have a hard time leaving the trap.
In his CloudWars newsletter, Bob Evans tracks the horse race among the largest cloud providers. In a recent issue he made the observation that “more and more businesspeople are coming to understand that ‘the cloud’ is becoming almost as much a mindset as it is a technology,” pointing out that ‘hybrid cloud’ means the seamless interconnection of on-premises systems.
Increasingly at CDS, we hear similar sentiments from both OEM partners and end-user customers. Our Raytrix MVS platform can extend the life of existing infrastructure so that data center leaders can buy the time needed to make sure the decisions between cloud and on-premises are made with all the available information, while enabling them to increase ROI and free up budgets for high-priority modernization and digital transformation projects.
There is no disputing the growth of the cloud model and the wealth it has created for the hyperscalers. According to CloudWars, the 10 largest cloud companies recently surpassed $7 trillion in combined market cap. That’s trillion with a T. Several of those companies are in other businesses (notably the top three, Microsoft, Amazon and Google) but Azure, AWS and Google Cloud are driving their valuations. If you are a company born during the last decade or so, whether to build out your capabilities in the cloud is rarely even a question.
But for the vast majority of the world’s large enterprises, cloud arrived after years and sometimes decades of careful infrastructure buildouts, talented technical staffs learning and perfecting customized apps, and continually more intensive security strategies. After years of too-good-to-be-true promises, the eyes are now wide open.