Data Centre ICT Cloud Storage Innovation of the Year
Founded in 2007, TeleData is an ISO27001 accredited data centre based in Manchester. It is the only data centre in the UK to offer an on-site NSI Gold Approved BS5979 security centre. TeleData offers colocation, cloud hosting, workplace recovery and data centre services to businesses across the UK.
In 2016 Teledata acquired cloud hosting company 1st Easy in order to expand our portfolio and capabilities. The intention was to extend 1st Easy’s cloud platform between multiple data centre locations, rather than relying on a single data centre, making a second site available for simple backup and recovery. This was to improve the resilience of the existing platform.
The team at TeleData then witnessed an incident where a close competitor suffered a single data centre outage. The impact on this provider’s customers was huge when they realised that their cloud hosted servers weren’t split between data centres for resilience. We decided to investigate further and looked into the marketplace to understand client perception vs reality when it comes to cloud expectations, and discovered that there was a real disconnect. Misconceptions and often a real lack of understanding surrounding cloud, was leaving cloud buyers unsure of what they were actually paying for, what the risks are and what questions they should be asking their providers.
Clients of the provider experiencing the outage simply weren’t aware of what they had - it wasn’t necessarily the providers’ fault, but the customers had assumed that as the provider had multiple data centres, then so did their cloud. Which begged the question, if the provider had multiple sites, why did customers not benefit from this?
We quickly decided that this wasn’t going to happen to us, or to our customers, so we not only accelerated our plans to make our cloud platform span data centres in true Active-Active configuration, but we also decided that we would migrate all our existing customers onto the platform, free of charge, to ensure that everybody would benefit from our investment.
Our team then went on to develop and launch CloudActiv - a cloud hosting platform built on VMWare with vSAN stretched cluster technology which spanned two Manchester data centres, with real Active-Active resilience for virtual infrastructures and networks. Using VMWare made the development of this platform easier than it may have otherwise been, but there was still a huge amount of work to do - especially to make the network perform how it needed to without impacting customer service, if one of the sites were to go down completely.
Being a colocation and data centre provider since the company started in 2007, we are well used to educating people about the underpinning nuts and bolts of hosting, so it was very surprising to us that cloud customers were not asking the same questions as physical hosting customers. Cloud customers were less inclined than colo customers, for example, to visit the data centres that would house their cloud data, inspect security and power protection systems, meet the people who would be looking after their business critical data, and spend time understanding any exposures that they should be aware of. In light of this, we set out to educate cloud buyers, and to deliver a cloud solution that provided - as standard - a solution that many customers blindly expected, but never received.
To add even more value to the customer and to help increase their efficiencies, we also released a CloudActiv self provisioning portal, which put control in the hands of the user, allowing them to add resources on the fly, without downtime.
As we had configured the entire platform to run with this Active-Active resilience, there was also zero configuration required by the client to make CloudActiv work as it needed to. Each and every virtual machine created was automatically deployed with resources available for every element to fail-over to a second site seamlessly.
It was a challenge to get all of our existing customers migrated from the old platform to the new, but with communication and planning, this process was extremely successful and we experienced zero issues with any individual clients.
One ongoing challenge of CloudActiv is capacity. To deliver site-to-site Active-Active cloud for every single element of resource, means we can never run either site, or any individual element of resource to more than 50% capacity to ensure that we can handle the sudden 100% fail-over from one to the other. This has major cost implications - both in terms of capital spend on hardware and recurring elements on power and rackspace.
By far the most complex element to get right was the network. We needed to remove any IP addressing concerns when advertising customer networks out of either of the two sites, which we managed to do with some expert configuration and design.
The result of all of this hard work, is a cloud hosting environment that is Active-Active, as standard - out of the box - whether for a single server or a complex IaaS environment. Disaster recovery is built in, there is no IP reconfiguration required in the event of a fail-over and an RTO/RPO of near zero. The project totalled a six figure investment for TeleData.
As well as increased resilience, the upgraded platform hardware has given us a huge amount more capacity and performance. Our monitoring systems tell us that the cloud platform is running quicker and more responsively than ever before, and we, and our customers now have an enormous amount of confidence in our critical infrastructure - confidence that could only previously have been achieved through customers spending a huge amount of money on secondary platforms, devops or a private cloud solution.