When an e-commerce date approaches, we take it for granted that many websites that are not prepared for such a flood of visits will crash. Amazon breaks its own sales records year after year whenever Black Friday, Cyber Monday or Prime Day arrives; on Black Friday, Worten’s web traffic is eight times that of a normal day; PcComponentes has a peak of six million visits; El Corte Inglés‘ online sales go up by 70% on sales promotion event days such as Black Friday; Renfe has suffered web collapses for months by quadrupling the capacity of its servers every time it launched promotions where more than 1,000 tickets were sold per minute.

Fortunately, error messages due to overloading have become less frequent (although not yet extinct). How has this been achieved? Well, through two desirable features in our digital solution: elasticity and scalability. And there are many mechanisms to achieve both objectives more or less effectively.

Any self-respecting website must be prepared to withstand a predictable barrage of requests, whether on a day like Black Friday or for a one-off promotion of its own. If, in addition to supporting these, our solution also solves unforeseen issues and is cost efficient, we would be taking advantage of one of the cloud’s most powerful features.


Elasticity: In physics, the term elasticity refers to the mechanical property of certain materials to undergo reversible deformation when they are subject to the action of external forces and to recover their original shape if these external forces are eliminated.

Scalability: It is an anglicism that describes the capacity of a business or system to grow in magnitude.

In digital terms, scalability is usually intrinsic to elasticity, since the method for achieving it involves the capacity to grow (or shrink) automatically under certain conditions, depending on demand. This is all very easy to achieve through cloud computing.

What do elasticity and scalability have to do with #BlackFriday?

What do elasticity and scalability have to do with #BlackFriday? Click To Tweet

The tools that help us ensure our solutions meet demand are: load balancers, performance metrics and automatic scaling groups.

Each cloud provider offers different alternatives but, in general, they are based on being able to calculate the capacity of a processing unit (normally our server) according to the demand it is capable of meeting and the associated mechanisms for generating more or less similar units and distributing the load between them. In this sense, while load balancers distribute traffic between our servers, the metrics add or remove new servers on the fly depending on the load metrics we are measuring.

You might think that this way of solving the problem is complex or expensive but, in reality, thanks to cloud services, these solutions are implemented on SaaS-type software or compositions in the foreground of our software. The same philosophy can be applied in a more or less sophisticated way to databases and other specific elements such as storage, buses, static distribution, caches. And as their cost grows proportionally to the traffic, this allows us to focus us on what we need to: our business and the conversion rate of those visits.

Solving the challenge of #BlackFriday and #PrimeDay or #CyberMonday in #ecommerce thanks to #cloud Click To Tweet

We might think that this is only useful if our software is custom-made in the cloud, but the industry is demanding solutions and there are also approaches to take advantage of these benefits using more traditional software.

Having software made of the same cloth as that used by large providers such as Amazon to meet the challenges of days such as Black Friday or Prime Day, is affordable and a requirement for the online presence of our business. In fact, AWS offers its own solution for e-commerce; this democratisation is part of the digital transformation and must be embraced if we are to compete in today’s cybersociety.

Image: pexels | rawpixel


  • Manuel de Paz

    Cloud Architect en Keepler. "Developer passionate about open source and cloud computing. I love programming, learning and sharing. My latest work is related to Microservices and Serverless, Hybrid Mobile App Development and Data Visualization".