By 2019, 83% of global data centre traffic will come from cloud services and applications, according to Cisco. The total amount of data centre traffic is expected to shoot from 3.4ZB to 10.4ZB between 2014 and 2019 respectively.
Eighteen percent, or 1.9ZB, of that traffic in three years time will be created between data centres and end users. 0.9ZB (9%) will come from data centres, such as replication and interdatabase links. The majority of traffic (73%; 7.6ZB) will be produced within the data centre (this includes storage, production, development and authentication traffic).
With the global data centre industry changing at pace, companies, consortiums, research organisations and others are putting together reports that look into this market to keep those involved up to date with the changes.
CBR rounds up five data centre reports launched this year.
CommScope: The connected and efficient data centre
Addressing the new factors that are increasing pressure on data centres (IoT, cloud, mobility and big data), telco company CommScope has produced a report around the "new holistic definition of efficiency" needed in the data centre space.
The document says that the majority of discussions on data centre efficiency focuses on power usage (for both power and cooling), however that only represents between 25% to 40% of a data centre’s monthly OpEx.
There is a whole 60% to 75% landscape of operating costs lying in other verticals such as space, IT assets, and connectivity. Therefore, a new holistic management strategy is needed through managing all of these physical resources as an interconnected whole.
Connected and efficient data centres can be managed as strategic business assets with financial metrics that transcend purchase price or utility bills.
New ways of measuring efficiency can be explored including availability, space utilization, workflow optimization, deferred and/or reduced CapEx, and more.
.
The report says that within the connected and efficient data centre, there is no shortage of opportunities to unleash the strategic value of physical infrastructure.
Veeam: Availability report 2016
Data centre disaster recovery and backup company Veeam has in February launched its latest annual edition of the "Availability report" that measures the gap between uptime and downtime of data centre applications.
After surveying 1,140 senior IT decision makers, the company found that the needs of the always-on enterprise are not being met, and that enterprises need to make availability a strategic priority or risk up to $16m each year in lost revenue.
24% of those surveyed ‘strongly agreed’ that despite all the investment in their organisation’s data centre, there is still an availability gap in terms of what IT can deliver and what users demand.
The report also found that businesses are experiencing 15 downtimes a year, with the average time of unplanned downtime for mission critical applications standing at 1.9 hours, and 5.8 hours for non-mission-critical applications.
The average recovery time objective (RTO) for mission-critical applications is 3.0 hours vs service level agreements (SLAs) requiring 1.6 hours. Similarly, current recovery point objective (RPO) is 4.2 hours, but should ideally be 2.9 hours.
Infinity: 5 reasons why CIOs need a new approach to data centre services
London data centre provider Infinity has found that one of the key areas that CIOs need to significantly re-evaluate is their approach to data centre services, because, firstly, the rate of change is unparalleled and unpredictable due to digital transformation.
Secondly, another area CIOs need to re-evaluate is the fact that the characteristics required of IT have changed and it is now required to be more flexible, dynamic and highly responsive.
Next, the company says data, and the insight it contains, is more valuable than ever, and that data has also never been more of a priority to enterprises than it is now.
CIOs also need to look at the fact that data centres are demanding a higher share of budgets, and Moore’s Law, which has been used to manage costs with hardware, software and end user costs, needs to be revised.
Lastly, businesses need more than ‘one size fits all’ in the sense that CIOs need to access to a range of flexible data centre solutions that reflect that, and when combined, can offer a tailored and optimal approach.
The Green Grid: EMEA research on data centre energy and resource efficiency
Global It and data centre consortium The Green Grid has unveiled another efficiency report, however, this time it exposes that the impact of data centres on corporate social responsibility (CSR) strategies is being hidden by the lack of clear energy efficiency definitions.
Of the 150 IT decision makers surveyed in the UK, France and Germany, 95% said that they are measuring efficiency. Yet, only 45% have configuration management in place, 57% defines energy efficiency objectives, and 85 % have a CSR in place implemented or in implementation.
Fifty-one percent of respondents believe board members see cost reduction as a top priority, which can be interpreted as resource efficiency and sustainability goals being farther down the scale.
Nearly 100% (97%) of people taking part in the study said that improvements in monitoring are needed.
The research has also found that virtualisation is the number one energy efficient strategy for most organisations (42%).
Ciena: Five data centre federation factors to help prepare for the IoT
With IoT spending set to boom from $1.4tr this year to $3tr by 2020, networking company Ciena has found that multi-tenant data centre (MTDC) providers are being challenged with balancing increasing IoT demands with those of the users connecting to content, and the enterprise customers who require enormous amounts of bandwidth and high performance.
The company says that a Data Centre Federation (DCF), could help manage all bandwith needs, "helping to keep all content and data consumers working, playing, and learning productively for years to come".
In order to realise the full benefits from the IoT, data centre providers need to address a data centre interconnect (DCI) infrastructure to start with, so data can travel through the network freely and securely.
Secondly, computational performance requirements are ever more required when it comes to data analytics, in order to gain meaningful, actionable insights from enormous amounts of structured and unstructured data. MTDC providers who invest in DCF will gain the ability to offer a consistent set of Service Level Agreements (SLAs) across all data centres within the federation.
Thirdly, for MTDC providers, the focus needs to shift from local storage to centralised and distributed storage capacity. They also must determine in what ways IT can harvest and use IoT data most cost-effectively. When cloud-based data is stored in an MTDC facility, DCF can be used to help drive low-latency, high-capacity performance.
Next, availability in the IoT becomes ever more critical, and MTDC operators need to address the need to design carrier-class networks, to support all appropriate SLA levels, from three to six-9s availability (9.99 to 9.99999%) or a maximum of approximately 30 seconds of downtime a year.
Lastly, Ciena says security is a key element and federated data centre network connections must be trusted, reliable, and secure, and aided by network encryption as warranted. Using DCF, MTDC providers can help enterprise customers encrypt all inter-data centre traffic to remove the possibility of intrusion.
