Managing traffic peaks with cloud caching
In digital transformation programmes, the number of website visitors has become a popular key performance indicator (KPI) for measuring brand engagement. But the power associated with a traffic spike can fast turn into a perilous “Sword of Damocles” if the website is not prepared. Cloud caching technologies that don’t require many costs upfront offer a solution.
In the world of e-commerce, website traffic is the modern version of ‘footfall’ and every company wants to get as much of it as possible. Indeed the number of site visitors is usually a top KPI. Given that, one would assume that websites would all be prepared for big hits of incoming traffic. Far from it!
Every year, recurring events such as big National Lottery jackpots, the Grand National races and Glastonbury ticket sales all lead to crashing sites and frustrated users venting their anger on social media and in the comment fields of online newspaper articles.
Unforeseen traffic spikes are even more complicated to handle. When the Duchess of Cambridge wore a dress from fashion retailer Reiss during a meeting with the Obamas in May 2011, Reiss’s website crashed for two and half hours.
Caching technology as a solution
Fortunately, there is a technical solution to ensure that websites perform, even at times of peak traffic. Foreseen or unforeseen. The magic solution is intelligent web caching.
A web cache contains copies of all the content of a website. It sits in front of a company’s backend server farm where it intercepts web requests and delivers copies of content to visitors. This means the company’s backend doesn’t have to reproduce multiple impressions of the same content, which can slow down or even lead to a website crash when many visitors arrive at the same time. Thus caching is critical for managing traffic spikes efficiently and it helps to improve website performance.
The alternative – adding infrastructure and caching servers to support traffic peaks that “might” happen is inefficient and expensive, potentially conflicting with other business KPIs to reduce cost and carbon footprint. Fortunately, most caching solutions can be used either on-premise or in the cloud making big infrastructure investments unnecessary.
Cloud caching = flexible web scaling
Cloud caching enables companies to choose when to scale capacity up or down depending on predicted traffic patterns such as last-minute bets at the Grand National. If unpredicted peaks occur, such as in the mentioned Reiss scenario, no hardware investments need to be made – companies just scale up and pay for the infrastructure they actually need and use during the peak.
However, there are scenarios for which a cloud-only deployment is not the right choice. One is when a company needs to maintain hands-on, local control for security or data-handling reasons. There are also certain operational tasks that require a more flexible approach. For example, an US-based global retailer might only have data centres in Nevada and California, but for regulatory compliance, reasons might be required to serve content from other locations in Europe.
In these scenarios, a company doesn’t need to renounce intelligent cloud caching completely but can opt for a hybrid deployment where the caching solution is installed on-premise and is complemented with the cloud version. The advantage of such an architecture is that the traffic can be dynamically directed between on-premise and the cloud, depending on the actual traffic needs.
If there is one thing that’s certain in our uncertain times it’s that traffic peaks can happen to any business at any time. Companies of all sizes require greater flexibility and more cost-efficient solutions to protect their sites, apps and CDNs while ensuring the best possible user experience. Deploying caching in the cloud or as a hybrid model is the most cost-efficient way to ensure that traffic spikes turn into revenue and loyalty, not a Sword of Damocles.
Hildur Smaradottir is Vice President of Global Marketing at Varnish Software, the company behind the open source HTTP engine Varnish Cache, used by more than 2.8 million websites worldwide. Hildur has 16 years experience in marketing, communication, and PR, and has worked 11 years in the tech industry. Hildur holds an MBA from Copenhagen Business School and a BA in psychology and journalism from the University of Iceland.