If we backtrack to 2016, we could find there were many large enterprises planning to transfer their datacenter workload to the cloud. In 2017, it is the year of transferring their business to the cloud, which causes a big stir in the whole data center industry, and brings great changes for cloud computing technology providers and data center providers.
Data center infrastructure provider Vertiv, formerly known as Emerson Network Power, has readjusted strategies, resources, and product combination to meet customers’ needs for transferring their business to its cloud infrastructure. Cloud computing providers start to develop their own hardware designs and supply chain function, so IT hardware providers have to struggle for surviving in the market. However, HPE decided to quit the competition among those cloud titans.
Those large-scale data center providers around the world have witnessed the unprecedented growth rate mostly driven by the cloud rental. Because companies are transferring more and more workloads to the cloud, cloud computing providers need larger data center space and capacity to support them, and many data center carriers like CyrusOne and Digital Reality Trust have benefited a lot from the trend.
In order to obtain more cloud computing business, data center providers have to largely increase their capacity, and drive the trend of integrating data center assets. It makes the number of those super-large-scale, independent data center carriers in the U.S be slimming down, so does in Europe and Asian-Pacific region. In below, we listed out some large data center acquisitions in 2017:
• Iron Mountain purchased IO;
• Equinix agreed to buy Metronode and will start market expansion in Australia;
• Peak 10 acquired ViaWest for $1.74B;
• Digital Reality bought Fabros;
• Digital Bridge purchased data center wholesaler Vantage;
• CenturyLink was acquired and turned into being Cyxtera.
Equinix acquired huge product combination from Verizon data centers in 2017, but it was released in December, 2017, so we didn’t list it above.
Although cloud computing business is increasing sharply, the world’s largest cloud computing providers now has admitted that those large companies they want to serve still have no plans to transfer all their applications to the cloud. Due to various reasons, these enterprises’ computational infrastructure will still reserve in infrastructure locally deployed or the data centers housing their servers. This fact becomes the focus in 2017. All cloud providers’ hybrid cloud strategy has begun to take shape, and these providers have different technology and partnerships with some enterprises and carriers.
Great Business Migration
In the early of this month, industry experts discussed with some operational staff from enterprise data centers and infrastructure at the Gartner IT and Operation Conference in Las Vegas. These operation employers said they are conducting a major cloud computing transfer project, or they are actively plan for the like programs. Most of them are working in world well-known companies.
To disclose the business relations is often forbidden by contract terms, but some of them are open to the plublic. Only in the fourth quarter of 2017, AWS announced it has reached significant cloud-related deals with GE, The Walt Disney Co., NFL, Turner, Toyota, and FICO. Microsoft recently made enterprise cloud deals with GE, UTC, BOA, Halliburton Company, and SBC.
According to their revenue increase rate reports, it is evident that enterprises are transferring to the cloud on a large scale. According to Synergy’s analysis on cloud computing providers’ Q3 revenues, the general income of cloud computing infrastructure services increased 40 percent per year. In marketing part, although AWS cloud computing still continues to expand, other companies’ market share grew rapidly as well. In fact, the cloud computing business income growth of Microsoft, Google and Alibaba is faster than AWS.
Adapt to the Fact of Hybrid Cloud
Cloud computing providers are managing to make customers use their on-site environment and cloud services. Then, hybrid cloud strategy became one of the biggest focuses in 2017.
In 2016, AWS and VMware announced partnership, and in 2017 they released their first set of products and services allowing customers to use VMware servers in specific AWS availability zones, and employ AWS Cloud platform to take backups and disaster recovery. Microsoft started to offer Azure Stack that can be integrated with public cloud. Google Cloud Platform announced a series of partnerships centering on hybrid cloud, and the collaborations with industry-leading hyper-converged infrastructure vendors Nutanix and Scale Computing were part of those. Meanwhile, Google and Cisco worked together, combining Cisco HyperFlex hardware stacks with Google software stacks Kubernetes for local deployment.
Cloud Computing is Extremely Flexible But Not Perfect
Many companies that have already run important applications on the cloud have learned a lesson in 2017 — companies’ applications can be run on the cloud, but to ensure the uptime still depends on the cloud itself.
In March 2017, an engineer entering a wrong order caused the service outage of Expedia, Coursera, Medium, Quora and Slack which are supported by AWS. According to an estimation made by a network risk modeling agency, the outage of AWS cloud services caused a loss of $150 million.
Experts said that AWS customers could cope with AWS cloud outage by investing extra time and fund in using more flexible models to build up their applications. In essence, customers can create redundancy across multiple availability zones, enable redundancy on multiple cloud platforms, or take advantage of caching service to save more than one copy in different locations, to address cloud service outage.
It is certain that super cloud computing platform has powerful resilience (it is even more resilient than many custom enterprise inner platforms). Companies operating these platforms employ top talents to guarantee business is running normally. However, these systems are not very reliable. As companies transfer more and more workloads to the cloud, they and cloud computing providers need together bear the responsibility for remaining normal uptime.