What is AWS? and why does good web hosting matter?
For a website to perform for an organisation and get results, it isn’t just design, functionality and user experience that matter. The reliability of the website itself in terms of being stable (remaining online) and the speed of the site both play a part in engendering customer loyalty and confidence. It only takes one visit to a broken or slow website to damage the perception of the organisation for the user. For high volume e-commerce sites an inaccessible website, even for a small amount of time, could mean a significant loss of earnings not to mention brand reputation.
Beyond the build of the application itself, it is the hosting and infrastructure around it that can we rely on to deliver the performance we need. So how do we as a digital agency create the optimal set-up for our customers? The needs of websites vary hugely and various things can have an impact, from the volume of traffic the site receives to the complexity of the functions and processes required. Hosting also varies hugely in scope and cost as well as in the levels of support and up-time that you can expect. Getting it right for customers used to be challenging, and traditional approaches simply required throwing money at the problem to get the best results.
Cloud hosting and Amazon Web Services
The development of cloud hosting platforms in recent years has changed the way we can now access and implement these services, revolutionising the digital landscape. At JUMP we have invested in Amazon Web Services (AWS), an extremely powerful service that has fundamentally changed the way we support our customers.
AWS is a toolkit of cloud-based services focussed on hosting, storage and content delivery, designed with a fresh approach to tackle many of the challenges that come with hosting a website.
AWS is a separate entity to Amazon’s online store, but that is where it has its roots. Over 10 years ago Amazon was forced to invest money in increased power to support the expected volume of traffic over the Christmas period. This was going to be an expensive outlay to serve a short time-period, but they were forced to go ahead to avoid losing sales and damaging their reputation. After the Christmas period they had a surplus of power which they proceeded to rent out to recoup their costs. So Amazon Web Services was born.
What are the benefits?
Amazon has a global infrastructure designed to allow organisations to service their customers around the world. The worldwide server locations and multiple servers within each region allow organisations to choose locations close to their customers for speed, but more crucially, this setup has been designed for optimum performance in the event of server outage. Their promise is to aim to provide 99.99 % uptime.
Reliability is not the only benefit of the AWS setup however; flexibility and scalability are also key: ten years ago it might have taken 6 months to increase the level of power to support a website, in recent years perhaps a few days, and now it can be done instantly. Traditional hosting companies allow us to purchase hosting of a fixed size for a fixed amount of time. The AWS Pay-as-you-go model allows its customers to rent space literally by the hour or even by the minute, and to change the amount of power flexibly.
As an illustration of the benefits: several years ago there were rumours that a high profile celebrity was going to mention one of our clients (a north-east based charity) on Twitter. We were concerned that their website, on a standard hosting package, might crash if the traffic suddenly increased, but we didn’t have the ability to respond to the circumstances, and didn’t want to tie the charity in to a costly package for a year. Had we had the flexibility of AWS, we could literally have doubled the hosting capacity for the following hours, bringing it back down again once the twitter frenzy died down, with no risk to the customer.
On a more recent project we were able to capitalise on the flexibility AWS has given us by renting a large amount of power for a short period of time to generate maps for a bespoke navigation application. Before the AWS model, the cost for renting the capacity we needed could have been prohibitive to the project’s success. Again Amazon’s initial problem has led to them creating something that has revolutionised the hosting provision.
Back to the Christmas rush problem that Amazon originally faced: Black Friday and Christmas sales have always been likely to strike fear into the hearts of developers, project managers and business-owners alike and no one wants to spend Christmas Eve tackling an offline website. AWS allows you to instigate pre-emptive scaling, allowing you to reserve increased space for specific periods of the year that will see heavy traffic, and allowing peace of mind that a website will be able to deal with these busy periods without it costing the earth.
Need for Speed
AWS was at the forefront of a wave of tech companies offering cloud web services such as Google Cloud, Azure and Digital Ocean. Despite the propensity of these services out there, AWS is still finding ways to stay ahead.
AWS has the perfect test-bed in Amazon’s online stores worldwide, and this often drives their innovation. They have recently invested in their own electricity and cabling infrastructure so they are not dependant on third party suppliers. This gives them dedicated internet cables to deliver top speeds, and allowing them to build new data centres and supply them with electricity in a matter of weeks.
They also take a different approach to storing an entire website on a server as one entity (as is the traditional approach), instead breaking it up into pieces (for example the website’s images on a specifically designed image server) to optimise performance and decrease risk of an entire website being offline.
Image credit: klevo / Shutterstock.com
JUMP’s developer dream-team
JUMP are part of the AWS Registered Partner programme and have members of the team specifically trained to build a robust infrastructure that suits our own needs. We have developed and standardised our infrastructure so that we can offer a secure and scalable service to our customers as reasonably as possibly. Standardising this infrastructure means that we can share resources across our customers and offer the benefits of AWS at a reasonable cost.
We’ve created a robust set-up which offers so much to our customers, but it’s by no means complete. We are continuously improving it and looking at ways to offer more. At JUMP we have a range of clients from simple brochure websites to bespoke online e-commerce platforms and it’s really exciting to be able to harness powerful tools to support these varying needs. We’re constantly looking at new ways to solve specific problems and when we find a good solution we’ll build that in to benefit future projects.
For example, we are currently exploring how AWS can help us improve the way a site processes email or sales requests to improve the customer experience and take pressure off the site during busy periods of activity. Each of these features take time to integrate into the infrastructure, but improves what we can offer and makes us more efficient.
We’ve also found that pairing AWS with other services is proving powerful, allowing us to tackle a whole host of recurring development challenges. We now automate processes that would have been prone to error, such as safeguarding against deploying a website which has problematic code.
Our end goal is to offer affordable, robust and high performing solutions to our customers. Any experienced development team will be familiar with the challenges that come with hosting, deploying and maintaining websites well and will appreciate how much value innovation in this area can bring to our industry. As expectations online continue to be raised, we want to be well positioned to ensure that neither JUMP nor our customers get left behind.