Valerie Potapova - Fotolia
Datacentre users could be paying over the odds for co-location services by insisting on running non-critical workloads in facilities with top levels of resiliency, it has been claimed.
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
Speaking to Computer Weekly at the IO Datacentre Showcase in Slough, Nigel Stevens, UK managing director of co-location provider IO, said buyers often insist on using Tier III facilities to run workloads that do not need such high levels of resiliency.
“Not everything needs to be run in a Tier III datacentre. In fact, I’d say at least 50% or more of an enterprise’s applications simply don’t need to be that highly resilient,” he said.
As enterprises become savvier about their datacentre needs, Stevens predicts resiliency will soon become a key area of focus for users looking to cut their datacentre costs.
“When customers start to realise that, it will make a material difference to their costs. For people who don’t need that level of resiliency, it could save them a huge amount of money,” he added.
The tier classification system was introduced by the Uptime Institute in the mid-nineties to provide a sliding scale of certifications that denote the resiliency and performance of datacentres.
The scale runs from Tier I to Tier IV, and to attain the latter requires operators to invest large sums of money in ensuring their infrastructure has sufficient redundancy and can deliver a service-level agreement (SLA) of at least 99.995%.
These highly available facilities are favoured by companies in the financial sector, for example, who use them for high-frequency trading apps and the like. However, not all companies will have apps and workloads with such high uptime requirements, said Stevens.
“It’s down to people identifying those things in their IT stack that don’t [need such high levels of availability], and seeing if there are cost savings to be made,” he said.
Read more about datacentre specifications
- The power usage effectiveness rating for measuring datacentre energy efficiency is often used by operators to differentiate their facilities from their competitors
- Research suggests inner cities are falling out of favour with datacentre developers, with cost-sensitive users willing to look further afield for co-location services
IO’s datacentres feature a modular design, which allows customers to physically lock their kit away inside a room that features an under-floor cooling system that provides the server on top of it with direct access to cold air.
The company’s Slough facility opened earlier in 2015 with Goldman Sachs as its anchor tenant, and has since attracted a number of other clients, said Stevens, including an insurance and a multi-national education-focused company.
As part of its work to entice more enterprises to use its Slough datacentre, IO has now partnered with the London Internet Exchange (LINX), meaning the facility now has a direct route to its network peering hubs.
This, said Stevens, will improve the levels of internet connectivity and latency its customers receive at its Slough datacentre.
“The proximity to internet peering this give us is the main benefit, and for anybody whose business requires them to be close to the internet, it’s a major plus point,” he added.