Picture credit: Diana Parkhouse/Flickr
The phenomenal growth in cloud based service delivery and managed services is changing essential data centre requirements. For any Managed Services or Software as a Service (SaaS) provider looking for a data centre to support this critical new business model and meet 99.999% customer Service Level Agreements (SLA), issues such as the resilience of connectivity and power, as well as security are clearly critical.
Organisations also need to consider the time it takes to get engineers on site – not only to address issues that may impact the SLA but also to set up and install new customer equipment quickly.
While every cloud clearly needs a home – it cannot be just any home.
New Imperative
Gartner predicts that by 2016 cloud computing will make up the bulk of new IT spend. 2016 is also predicted to be a defining year as private cloud begins to give way to hybrid cloud and nearly half of large enterprises are expected to have hybrid cloud deployments by the end of 2017. The resultant shift towards Software-as-a-Service (SaaS), Platform-as-a-Service (PaaS) and Infrastructure-as-a-Service (IaaS) is having a fundamental impact on data centre priorities.
These cloud services require a physical infrastructure: a safe and secure environment from which the business critical servers and hard drives that contain the applications and data can operate. They require fast, uninterrupted data connections, a totally secure and reliable power supply and buildings that offer the highest possible levels of security.
Five keys to finding the right home for the cloud
Key 1: Location
The right data centre location is not just about the obvious requirements of avoiding the flood plain or flight path, although these are clearly essential. For any Managed Services provider there are a number of other considerations, not least the time it takes engineers to travel to and from the data centre. While travel time is less of an issue for a colocation site that will rely on remote management and data centre staff for the majority of support requirements, most Managed Services providers have a far more hands on approach.
From adding new customer equipment, to the need to ensure any issues are immediately addressed to meet customer Service Level Agreements (SLA) and avoid problems, the speed with which engineers can reach the data centre can be critical. A data centre in central London that can be accessed in minutes rather than hours reduces wasted travel time and costs, enabling a far more responsive service.
Key 2: Business Scalability
With many Managed Services providers making their first foray into the cloud, it is extremely tough to predict business growth – and hence data centre requirements. It is therefore essential to look for an environment that offers flexibility. For example, can the data centre run high density and low density racks side by side? Without row-based, rack level cooling, an organisation might struggle to adapt or add racks without the upheaval of moving to a different location within the building. Checking the options available for adding and expanding racks is essential for any provider expecting to rapidly scale up its cloud based business.
Key 3: Connectivity
Cisco estimates that by 2017 global cloud traffic will reach 5.3 zettabytes (one zettabyte of data is the equivalent to the information stored on around 250 billion DVDs). In this environment the issue is not only the reliability of the connectivity but also latency: the ability to offer transmission time in nanoseconds fundamentally transforms the quality and type of solutions and services that can be offered.
In terms of resilience, every data centre offers multiple Tier 2 carrier options. Few, however, are able to offer multiple Tier 1 carriers. As a result, this means that the connectivity is still reliant upon a single fibre provider – typically BT. Without diverse fibre connections, the back-up options are limited: any damage to the underlying cable network – for example, damage during road works – will take out every Tier 2 connection that is simply using the same last mile from BT. For any Managed Services provider it is essential to look for a data centre with multiple Tier 1 carrier options and diverse entry points into the building as well as Tier 2 carrier services, in order to achieve full resilience.
Key 4: Power Supply
All Data Centres can be expected to have a dual power supply from the grid; and all have some form of resiliency or back up to the grid power supply from an array of batteries or generators. How many in central London, however, can offer dedicated 33KV transformers? Most urban data centres are limited to 11KV tapped into one main grid substation – if that goes out, they will lose both A and B grid feeds. A data centre should take dual feeds from different substations to provide true resilience – however, given the limitations regarding power supply in London this is becoming harder and harder to achieve creating potentially huge risks for Managed Services providers.
Key 5: Security
Security has always been a vital consideration within any data centre environment. From a bomb proof building to multiple levels of security to access the building and the server room and the additional option of adding cages around the racks, or even a dedicated private suite, a depth of security facilities is key to support the changing threat risk. Given customer concerns regarding the data security in the cloud, it is essential for any Managed Services provider to be highly confident in the procedures in place, from authorised and unauthorised access to fire and vandalism risks.
Conclusion
The cloud has transformed the way businesses operate and the way the world does business. However, in what is becoming a commodity marketplace there is a risk that organisations are failing to understand the essential data centre requirements. This is a business critical decision – rather than buying down to a price, expectations should be rising and organisations should be continually looking for more from providers to support an evolving business model.
The right environment does not only minimise risk but also reduces costs by cutting non-productive engineer time and improves responsiveness to customers. Isn’t it time to unlock the right home for the cloud?
