un@frust.co🇨🇱 Callao 2911, of 4144, Santiago, RM, 7550285🇺🇸 1111B S Governors Ave STE 29963, Dover, DE 19904
A fictional story based on real data
Cyber Day is not just a sales event — it is one of the biggest stress tests for company technology infrastructure. In this post we analyze what happened to our clients' AWS infrastructure during the three-day event (June 1–3) to give you a picture of what may have happened with your own.
More than 70% of our clients recorded a significant increase in activity compared to the week before Cyber Day. This includes companies that do not sell directly to end consumers: payment platforms, logistics systems, admin applications, data integrations.
The effect is clear: Cyber Day is not just a retail event. It is an event for the entire digital economy — and infrastructure reflected that precisely.
CDN requests grew more than 43% over the three days of Cyber Day — from an average of 375 million to 539 million requests.
Edge compute services (personalization, routing, authentication) nearly doubled (+93%) compared to the prior week.
Infrastructure spend grew more than double during the event (at the median). One client reached nearly 13× their usual consumption on the peak day. No manual intervention was needed: no emergency calls, no last-minute upgrades.
What's interesting is not just how much it grew — but how. Traditional servers (EC2) barely moved (less than 5%), because that capacity was already reserved and running before the event. Useful, but expensive: you pay for it 365 days a year.
Container-based services grew more than 22% — they spun up during demand spikes and came back down afterward. It's the difference between hiring permanent staff for a one-time event versus calling in reinforcements only when you need them. The business paid for extra capacity exactly during the days it needed it, and not a day more.
If there is one conclusion every CTO should take from this analysis: the bottleneck in a high-demand event is not the servers — it's the data layer.
Cache and database services grew far beyond the rest of infrastructure:
An additional server can be up in seconds. A relational database cannot scale the same way — it requires advance planning, read replicas, caching strategies. Clients who arrived at Cyber Day with that layer well-prepared absorbed the load without issues. Those who did not, likely felt it.
Comparing clients with container-based architectures versus those without, the difference is concrete:
Clients with more modern architectures grew 50% more than the rest.
The business read is direct: investing in modernizing architecture is not just a technical decision. It determines whether the business can grow when it matters most — or whether technology becomes the ceiling that limits sales.
During Cyber Day, the volume of activity monitored by security systems grew more than 60%. This includes transactions, data access, API calls — all analyzed in real time by Amazon GuardDuty, active in 65% of our clients.
Cyber Day 2026 was, from an infrastructure standpoint, a successful event for most companies that arrived prepared. The cloud did what it promises: scale when needed, without human intervention, without downtime.
The numbers left on the table for next year are those of databases and architectures that have not yet transitioned to more elastic models. That is where the difference between being prepared and not is measured in lost sales during peak hours.
At Frust we work with technical teams to understand how optimized their infrastructure really is — and free up budget so it can be directed toward improvement and modernization projects.
Veja quanto sua empresa poderia economizar na AWS
Conecte sua conta e receba uma estimativa de economia em menos de 48 horas. Sem compromisso.
