IT specialists state that nowadays, Hyperconvergent infrastructure solutions market is ranked to be one of the most dynamic segments of the IT industry, that will encounter a tremendous development in the next 4-5 years.
According to Technology Business Research (TBR) estimated analysis, the annual growth rate for the next four years will be about 50%, generating by 2020 – when nearly a third of Data Center infrastructures will be hyperconvergent – a total value of 1,6 billion USD. The pace of adoption will be accelerating in the coming period, TBR expecting that hyperconvergence will become in 2025 a major technology in datacenter hubs.
With a more optimistic perspective on the evolution of market values, Gartner has estimated that in three years by now, the value of this industry will be 5 billion USD. The prediction was partially confirmed by International Data Corporation (IDC), estimating a 2 billion USD worth this year that would double by 2019 (3.9 billion dollars). IDC analysis is based on the results from the previous years. In 2014, the market on hyperconvergent infrastructure solutions increased by 162.3 %, reaching a total value of 373 million USD, and last year registered a 116 % growth, reaching 807 million USD.
The spectacular developments expected by analysts are caused by the accelerated migration of the organizations towards hyperconvergent infrastructures and software-defined technologies. It is a phenomenon generated by the fact that more and more data centers are conditioned by out of date infrastructure models, which fail to cover effectively the demands of new type of distributed applications with processing needs and increased IOPS, increasing the level of virtualization and cloud adoption extensive services.
According to statistic studies, currently, a classic Data Center infrastructure includes hardware and software products from 8-12 different vendors, each with its own management interface. This requires an IT team with a vast array of skills, able to cover those technologies, especially to identify and solve in a timely manner, the incompatibility and integration problems.
Another major problem of the classical model is that it no longer can provide the effective use of resources – usually every equipment is overprovisioned to handle peak loads that occur more or less randomly. The result – a waste of processing capacity, memory, storage and network, that cannot be reallocated dynamically only by a costly effort for setup and administration.
On the other hand, the trials of solving these types of challenges throughout the virtualization solutions don’t manage to solve the problem entirely. Because for the majority of servers the virtualization projects are limited, this leads to over-provisioned “islands”, that don’t succeed to split the resources between them.
A first response of the IT industry to this amount of challenges was represented by the integrated systems, considered to be the predecessors of today’s convergent infrastructure solutions. The answer – although it covers the processing requirements area, storage and network, integrating virtualization functionalities – didn’t succeed to offer the also desired functionality and scalability, thus being an aggregation of hardware and software products with a not too advanced capability. Moreover, integrated systems have generated other two major problems – they raised the client’s dependence on technology vendor (“vendor lock-in” effect) and implied substantial investments when a resources’ supply was necessary (meaning the acquisition of an entire equipment).
The second type of answer is represented by convergent infrastructure solutions that integrate server and storage components in only one equipment, eliminating the need for dedicated storage equipment. This kind of systems have a verified compatibility level and facilitates, through virtualization, access to only one collector of resources, which is easy to manage and allows fast developments. From the integrated components perspective, the total acquisition cost is low, and the management (at server and storage level) is simplified. The most frequently used example in explaining the benefits of converged infrastructure solutions is Vblock System, the product resulted between VMware-Cisco-EMC collaboration (marketed under VCE acronym). Vblock integrates VMware vSphere platform across UCS servers and Cisco network solutions, plus storage systems from EMC. The system has a high compatibility level, by not only providing support for VMware virtualization but also for KVM and Hyper-V. The product is designed so as to allow the creation of a modular architecture by adding blocks depending on the arising needs. The systems are pre-tested and are delivered pre-configured, which considerably reduces the time of commissioning and development effort. The reputation of the Vblock systems stands out because, according to Gartner, VCE is a leading converged infrastructure solutions. However, converged infrastructure solutions have some important drawbacks. The first would be that the ratio resources CPU / storage/network is fixed, which is not entirely answering the flexibility needs of organizations. The second drawback is that the solution cannot always be used together with existing data center infrastructure, namely “legacy” systems. This leads again to creating islands of resources and does not solve the performance problems of old infrastructures.
Chronologically, hyperconvergence is the third type of response offered by the IT industry. A solution with extensive coverage, as demonstrated by market figures.
Hyperconvergence infrastructure solutions manage to eliminate the convergence limitations, providing a major simplification by consolidating the entire infrastructure into a single reservoir of resources that can be dynamically allocated automatically. Effort management is substantially reduced and savings are realized both in the operational area (eliminating islands of technology leads to the disappearance of dedicated teams) and the level of investments ( CAPEX).
Hyperconvergence arguments are:
- Increased flexibility – adding/removing of resources can be achieved quickly, depending on the evolution of business requirements;
- Improved levels of availability and data protection – hyperconvergent infrastructure solutions integrate automatic restore features that prevent loss and/or data damage, reducing the risk of downtime;
- Efficient allocation of resources – through automatic and dynamic allocation of storage space, processing power and bandwidth depending on load peaks and specific requirements of each type of applications and/or business process;
- Simplifying virtualization processes – hyperconvergence infrastructure solutions are focused on virtualization, enabling rapid development of virtual machines and balancing loads;
- Reduced costs – by eliminating areas of incompatibility and simplify resource sharing, hyperconvergence provide substantial savings that can be tested quickly.
Hyperconvergence – the only reservoir of automatic resources allocated dynamically
All these arguments are not only to promote vendors’ messages but confirmed in practice. ActualTech Media released a study last year, that analyzed the benefits expected by companies that intend to adopt hyperconvergent infrastructure solutions by reference to the benefits of the organizations that have already implemented them. The differences between the IT departments’ expectations and the results are minimal:
- Improving the level of scalability: a benefit monitored by 48% of companies and obtained by 42% of organizations that have adopted hyperconvergence;
- Increase in operational efficiency: 43% expected, 37% obtained;
- Reduced development time: expected 34%, 32% achieved;
- Cost Savings: 34% expected, 31% obtained;
- Streamline backup and restore processes: expected 28%, 28% earned;
- Increase of agility in provisioning virtual machines: expected 23%, 23% obtained.
The results presented show that organizations properly understand the advantages and gains offered by hyperconvergence, which explains the major existing interest in this area and justifies predictions of rapid growth in the rate of adoption.
Hyperconvergence is approached dually on the market: on the one hand, there are software solutions vendors camp – Nutanix, SimpliVity, Pivot3, SpringPath, Gridstore, etc. – and on the other hand, the OEM solutions manufacturers represented on the top by Cisco and HP. A rigorous separation is still difficult to make because new competitors appear constantly (according to IDC, at the end of 2014, about 20 companies had announced plans to launch products and/or hyperconvergent solutions) and new partnerships. For example, only earlier this year, Cisco announced partnerships with SimpliVity and SpringPath.
Cisco has made another important move by launching in early March c.a. (at Cisco Partner Summit) the HyperFlex Systems (HX Data Platform), which analysts quote it as a future market leader product. The development of HyperFlex hyperconvergent infrastructure solution is based upon UCS Server components (product leader in the US market of x86 blade servers and number two worldwide) Cisco Fabric Interconnects network solution (10 Gigabit Ethernet), integrates HALO solution from Springpath and provides support for VMware hypervisors. One of the main competitive advantages of the new Cisco product is that it offers a high level of flexibility and scalability, the addition of resources on the storage or processing area wouldn’t require the acquisition of new equipment but adds in the chassis of the required units. Designed on the plug-and-play system, HyperFlex can be configured and scaled quickly, reducing the deployment time to several minutes. HX platform integrates advanced deduplication and compression features inline for reducing data volumes by 80%, and solutions cloning, snapshots, and restoration, that increase data availability. Cisco product is designed so as to ensure a reduction of up to 30% of the total cost of ownership and an extra 40% performance (compared to competing solutions). Positioned under the “Data Center in a Box” name, currently, Cisco HyperFlex system comes in three configuration versions, providing a simplified licensing system.