Don’t forget, it’s the hardware that makes the cloud
The main issues we see with clients and cloud implementations are that it can be very difficult for them to get a clear idea of what it is they are buying and how well it will perform.
While the consumption and billing models are clear, it can still be hard to know how much you will pay each month. But what is hard is predicting exactly what the level of performance you will get.
Some of this is inevitable. In the end we all rely on a length of copper wire, or fibre-optic cable if we’re lucky, to supply the final link between the cloud and our home or work computer. This critical part can be the downfall to an otherwise perfect set-up.
But it can also be very difficult to compare different cloud providers – or understand the different services offered by the same provider. It is worth remembering that what you’re paying for is access to someone else’s hardware, something that comes with its own limitations. And you also need to consider the different stages of network access that you’ll need to access that hardware.
Shifting clouds need strategic thinking
Over the many years that we’ve been helping clients build successful, high-performance cloud infrastructures, we have seen a shift in how the cloud is used and what people’s expectations of it are. No longer is it just for back-up or to process overnight batch work.
Business clients expect cloud services to be as responsive, and to work as quickly, as their applications running locally. And that is only possible with a deep dive into each networking and inter-connection across your entire estate.
There is a general misconception that somehow these challenges have gone away. At Pulsant, we’ve invested £8m in a state-of-the-art network supplying fast and affordable links between all our distributed data centres and delivering sub-5 millisecond latency to 95% of the UK population. That means we can take your workload or application and put it where it needs to be. That might be close to your head office or near to one of your clients. It could be very close to a specific data centre or hyperscaler.
Getting the very best from your cloud investment requires this sort of granular, geographic detail. We often help clients struggling with latency and resilience issues who find their data is being sent around the country, often via London, instead of via a more direct route. We can also help with moving workloads throughout the working day as demand changes.
Hybrid networks need detailed planning
This affects even the most advanced clients. The gaming industry for instance is hardly a technology laggard. But even games developers usually rely on a complex hybrid infrastructure to support their on-premises hardware.
The cloud has not done away with network topology – it has just made it more complicated.
The most famous example of this is the Amazon Web Services (AWS) Snowmobile. Crazy as it sounds, this is a 45-foot-long shipping container towed behind a lorry. In some cases, the best way for AWS to shift serious amounts of data is by road. The lorry arrives at the data centre, plugs into the network, and loads up to 100 petabytes of data. Then it drives to the AWS data centre it needs to get to and downloads its cargo.
Now the chances are that you don’t need a lorry and a team of security guards to move your data around. But you do need to think strategically about how you can make the most out of your cloud usage – you can’t just ‘put it in the cloud’ and forget about it.
At Pulsant we have years of experience of building data centres and networks that do exactly what our clients want and need. We can help you dig down into workloads, hardware, connectivity, and interconnections to make better strategic decisions about building and developing your hybrid cloud implementation. Taking the time to consider what you actually need out of the cloud is your number one priority. It means you won’t take your IT problems with you or create new ones.