From the desk of Leon Henery, CTO, AFD Software
When outages strike, continuity counts
October’s cloud failures were a masterclass in fragility. A major Amazon Web Services (AWS) region collapsed from a Domain Name System (DNS – the system that translates website names into the actual servers they run on) defect. Azure’s Content Delivery Network (CDN – a system that helps deliver online content quickly) misconfiguration followed nine days later. Between them, they knocked thousands of organisations offline.
The UK escaped relatively unscathed, but that’s cold comfort. The technical failures that crippled Virginia could just as easily strike London tomorrow. At AFD Software, we’ve never designed our architecture around hope.
Beyond the Myth of the Infallible Cloud
The AWS outage on 20th October started with a latent defect in US-EAST-1’s DNS infrastructure. With no way to translate a website URL to the IP address, within minutes, applications couldn’t resolve service endpoints, triggering cascading failures. Major brands lost access to critical systems for hours.
We’ve never accepted the assumption that cloud giants are infallible. Our approach has been consistent: every system can fail, so design for resilience. That principle drives the architecture behind our Evolution Platform, the engine that powers all our services.
Built for Continuity
For more than a decade, AFD has operated two independent, production-grade environments: one in AWS, providing a high performance cloud platform across multiple availability zones; the other in our own Tier 3-equivalent data centre at our Isle of Man head office, with independent power, connectivity, and physical security.
These aren’t just backups, both are live, always ready to take 100% of customer load. This active-active design means zero-downtime maintenance, real geographic diversity, and no single point of failure. On a product level, calls that fail in AWS are automatically routed to our own data centre, and if problems persist, all traffic can be redirected in under 60 seconds.


Unique: Extended Resilience
We extend this resilience to our customers. Beyond our hosted solution, AFD offers on-premises deployment giving organisations complete control. Windows installers or Docker images are available to be installed on your site and can operate standalone or as part of a hybrid architecture. This hybrid approach provides control where you need it, with the option to
failover to AFD’s hosted infrastructure when your systems face unexpected disruption. Whether fully hosted, on-prem, or hybrid, AFD ensures your services remain available.
Measured in Minutes, Proven in Years
Numbers tell the story best. Over the past four years, AFD has maintained uptime exceeding 99.99% across our hosted platforms, surpassing even our own internal targets. In practice, that represents only seconds of potential downtime each year, with no full-service interruptions across our combined Tier 3 Data Centre at our Mountain View HQ and AWS environments.
That level of reliability doesn’t happen by accident. It’s the result of continuous investment and a culture that treats uptime not as a statistic, but as a promise.
Team and Technology
At the heart of every resilient system is a resilient team. Our DevOps team continually refines our infrastructure for security, speed, and efficiency. They ensure that AWS Shield defends against Distributed Denial of Service attacks, our Web Application Firewall filters malicious traffic, and our load balancers direct requests intelligently to healthy nodes.
If any single environment is compromised, we are ready. In the extreme event that both environments were simultaneously compromised it takes just 45 minutes to create and deploy a completely new healthy production environment. That is one benefit of our truly agnostic, fully automated architecture. But our greatest strength isn’t automation, it’s our attitude. We don’t accept outages. We engineer around them.
Eyes on the Horizon
Cloud computing remains indispensable; its scale and elasticity make it hard to ignore. But October’s failures proved that even giants aren’t infallible.
The question isn’t whether to use the cloud, but whether you can rely on it alone. At AFD, we give customers architectural choice without compromise: deploy our products in your data centre, use our dual-region hosted environment, or build a hybrid solution that fails over seamlessly between your systems and ours.
Every address lookup, every payment, every moment of trust matters. We care about your business, your growth, your success. That’s why at AFD, reliability isn’t just claimed – it’s built.
To discuss how AFD Software can help strengthen your organisation’s resilience strategy,
contact our team on 0333 433 0711 or email postcode@afd.co.uk.
We use cookies to enhance your browsing experience, and analyse our traffic. By clicking "Accept", you consent to our use of cookies.