Website failure is all too common. Some of the biggest companies in the world have experienced it — think Netflix and MasterCard — and while many bounce back relatively unscathed, for others it can prompt a customer exodus, or even lead to the death of a business.
For startups, the risk increases exponentially. And given numerous reports suggest 50 percent of consumers will abandon a website if the page does not load within three seconds, with 80 percent of those people not returning, it is easy to see how critical it is for startups to have the right infrastructure to ensure near-100 percent uptime.
Website failures can occur for a number of reasons: Distributed Denial of Service (DDoS) attacks, DNS problems, and poor quality web hosting are all common issues. So too are traffic spikes, which occur when a web hosting environment is overwhelmed by the number of users trying to access a page.
While most startups can’t imagine this impacting them, the truth is all websites are susceptible to spikes. Spikes can be great news — it means thousands of potential customers are checking you out — but if not handled correctly, that traffic is harmful more than it is beneficial. So, what do startups need to consider to avoid a website meltdown?
Selecting the right web hosting environment
Your web hosting environment is one of the most important elements to your online business, so deciding whether to use more traditional hosting services or the cloud is a critical first step. Cloud services, including Amazon Web Services (AWS) or Microsoft’s Windows Azure, are becoming increasingly popular choices for startups, offering scalability, high performance, and availability, as well as robust security for growing businesses.
Importantly, it also eliminates single points of failure. Managed across multiple servers in different geographical locations, traffic can be automatically redirected to a different server if the first becomes inaccessible.
However, the cloud by itself isn’t a salve. For example, many of AWS’ most fervent customers have suffered outages when a single region has failed, because their servers weren’t configured properly to negate this potential issue. Architecting your cloud setup so these large scale failures don’t affect your business may not be obvious but is completely critical.
Ultimately, the decision depends on your business needs — both now and in the future. While traditional hosting can be a cost-effective solution for startups in the interim, the cloud ensures your hosting platform grows along with your business.
Once you have selected your hosting platform you need to develop a load balancing strategy. Essentially, load balancing will help you distribute incoming web traffic among servers hosting your content to prevent any one server from becoming a single point of failure, ultimately improving overall application availability, responsiveness, and scalability.
It’s also an easier way to scale out your infrastructure. As demand increases, new servers can be easily added to the resource pool so the load balancer will immediately begin sending traffic to the new server.
Just setting these up isn’t enough — testing your load balancing capabilities determines what your platform can and cannot handle. The best time to perform these tests is before peak traffic seasons, such as holidays and during new ad campaigns. That way you can make sure your site can handle the largest number of visitors.
Auto Scaling is another critical factor that will help your website seamlessly withstand almost any traffic demand, whether planned or unplanned. It works by maintaining application availability and allows you to scale your hosting environment capacity up or down automatically according to the conditions you set. This is particularly useful during spikes in demand to maintain performance. It also decreases capacity during low times to reduce costs – particularly valuable to startups.
Sseko Designs, an e-commerce brand that employs women in Uganda to make sandals that help provide them with a means to pay for their college education, is a perfect example of how auto-scaling can help businesses withstand traffic spikes without spending excess money on unnecessary infrastructure.
The company recently secured a rare spot on hit television show Shark Tank in the US. Based on viewership statistics for the show, Sseko expected its appearance to lead to 30 times more traffic than its existing Magento site was designed to handle.
Understanding the ramifications of a site meltdown, it opted for an AWS environment, which would enable it to leverage its mission-critical auto-scaling configuration setup. With the AWS environment, Sseko could scale its system up an hour before its Shark Tank appearance and 30 minutes after the show, scale back down, making it a cost effective solution for the startup.
On the night of the first Shark Tank appearance, Sseko’s website recorded 45,000 users with a 2,577 percent year-on-year increase for the week following the show. Revenue also shot up 1,313 percent compared to the same period the year before. Shortly after, Sseko Designs also featured on Good Morning America, during which it was processing 10 orders a minute and had a 908 percent increase in site visitors.
This example shows just how important auto-scaling capabilities can be for your online business. Regardless of whether you typically experience a relatively stable demand pattern, regular fluctuations in traffic, or one-off spikes, auto scaling ensures always-on availability without breaking the bank.
Techniques to prevent meltdowns
When it comes to handling traffic spikes, there are a few simple techniques businesses can use to avoid falling victim to website meltdowns. Make sure you choose a hosting environment that fits the need of your business now and in the future, be aware of what your site can and can’t handle, and ensure you have the capabilities to scale up and down quickly and easily depending on traffic demands. It’s these little things that can really help your business survive any traffic condition.