Does your website rely on viral traffic?
Has your site been brought down due to traffic spikes – from both legitimate viral hits, as well as unwanted denial-of-service attacks?
Facebook and Twitter can be good sources of traffic. With the right combination of headlines, images and timing, you can get that viral hit, which can hopefully either jumpstart return traffic or at least catapult your brand into the collective consciousness of social media users. But it’s not enough that you can bring in millions of readers. You also need the means to support huge increases in traffic.
In this article, we will share how a high availability infrastructure can ensure your website does not go down because of traffic spikes.
The Viral Hit
Viral marketing is not exactly a new concept, but using the Internet as a medium of distribution has made it an especially effective means of reaching out, especially if you get to pique user interest correctly. From supposedly found-footage format horror movies to blending iPhones, brand marketers want to find the sweet spot in engaging an audience well enough for their message to go viral. Every content producer dreams of that viral hit that would rake in a million page views in a day or a week. However, building demand is one thing. Actually being able to serve that is another.
Such authors or publishers may be merely focusing on creating potentially viral content without thinking about what could happen to their servers when all that traffic really comes in at such short notice. What’s the use of having a really awesome video of a cat singing “My Way” if people cannot access it when they click? There goes your chance at getting a viral hit!
How Not to Handle a Viral Hit
During the 2013 Super Bowl, the Coca Cola Company ran an interactive ad that directed people to a special mini site that would allow them to vote for the ad’s ending. Their server could not take the amount of traffic that came in, and the site had a load time of a little more than a minute. The server soon crashed.
In effect, rather than going viral, the company paid millions of dollars just to piss their customers off. And Coke wasn’t alone that night. SodaStream, Calvin Klein, Axe, and the Walking Dead all ran ads for the night and soon had their websites out of commission.
Improving Your Infrastructure
So how do you ensure that your server can take the traffic spikes that come with a viral hit? A popular workaround is to implement a CDN or caching mechanisms. This involves serving up static content instead of directing traffic to the database every time someone accesses your content. This way, your users would still be able to access the content without having to tax your database servers too much. A CDN will also deliver content from the server closest to the user, potentially speeding up load times.
Of course one of the better ways of ensuring availability is deploying your website on top of a scalable infrastructure, such as a cloud hosting solution. Rather than rely on a single point of access – which can also be a single point of failure – you get a more elastic service, which can grow or shrink depending on need.
The best thing that a publisher can do is incorporate high availability (HA) management you’re your infrastructure to make sure that you get to avoid crashes and get reliability benefits along the way.
High Availability Management
Having a content delivery network for your static assets is good if your website experiences traffic spikes every now and then. But why not make it better? A high availability infrastructure can also give you the capability to monitor your traffic, balance the load across servers to prevent overloads, thus routing traffic and application requests to the most viable ones. You also get a disaster recovery platform that can get your service back up and running in the event of a crash.
A good high availability platform can also include features that block unwanted traffic, thus protecting your website from distributed denial of service (DDoS) attacks and other hacking attempts, using advanced cloud-based firewalls and traffic algorithms.
Effectively, this does not only involve relying on caches to direct traffic away from your live servers, but you gain the following benefits.
- Ensure that you manage and direct all incoming traffic to the servers with best availability.
- Do more with your content with analytics, which can help you optimize content for better conversion (such as subscriptions or sales).
- Protect your website from hacking and other attacks that can take it down.
Benefits Outweigh the Costs
Social media platforms like Facebook, Twitter, Linkedin and others, employ high availability management in ensuring their infrastructure can adequately service hundreds of millions of users concurrently. Facebook’s Autoscale load balancing system even comes with the added benefit of improved energy efficiency. Because traffic is balanced appropriately across servers depending on utilization, demand and other factors Facebook’s datacenters run cooler and save on electricity usage.
Traditionally, load balancing appliances used to make a significant dent in the business budget. With SaaS-based services, you can start small and then scale along the way, only paying for actual usage. Being cloud services, these do not require hardware, personnel and maintenance costs. You can even be billed according to utilization or time, which can be particularly economical — after all, traffic surges don’t happen every day.
High availability, or being online for 99.99% of the time, has its benefits. Protecting your brand reputation, giving your customers the access they want, and not wasting money as Coke did, will certainly make the subscription prices for these services just worth it.