In the space of just ten days, both Microsoft Azure and Amazon Web Services (AWS) suffered major outages that disrupted thousands of businesses and millions of users worldwide. From government portals and financial institutions to airlines and retail giants, the impact was immediate and far-reaching. These incidents have reignited a critical debate: are we too dependent on too few cloud providers? And more importantly, what can organisations do to build true resilience into their digital infrastructure?
What Happened
On October 20th, AWS experienced a 15-hour outage triggered by a DNS failure in its US-EAST-1 region. The fault cascaded through core services like DynamoDB, EC2, and Lambda, leaving thousands of companies offline and millions of users locked out of essential services. Just days later, Azure suffered a similar disruption due to a misconfiguration in its global content delivery network, affecting Microsoft 365, Xbox, and a host of enterprise applications.
These weren’t isolated glitches. They were systemic failures that exposed the fragility of centralised cloud infrastructure. Despite being designed for high availability, both platforms revealed how a single point of failure can ripple across industries and geographies.
What is DNS?
The Domain Name System (DNS) is the Internet’s “phone book,” translating human-readable domain names (such as google.com) into the machine-readable numerical IP addresses that computers use to locate and communicate with one another.
Why DNS Matters
DNS is a critical and foundational component of the internet’s infrastructure for several key reasons:
- Ease of Use: It allows people to use memorable domain names instead of having to recall complex strings of IP addresses (e.g., 142.250.191.14).
- Essential Functionality: Nearly every internet service, including web browsing, email, and cloud applications, relies on DNS to function correctly. If DNS fails, these services become unreachable.
- Flexibility and Scalability: The distributed and hierarchical nature of DNS allows website owners to change the physical location (IP address) of their services without affecting end-users, who continue to use the same domain name. It also ensures the system can handle the vast number of domain names and requests globally.
- Reliability and Performance: DNS uses a distributed network with multiple redundant servers and caching to provide fast and fault-tolerant service.
How It Helped Cause Recent Outages
DNS issues were the primary cause of several recent widespread internet outages, most notably those affecting Amazon Web Services (AWS) and Microsoft Azure. The outages were generally not due to a fundamental failure of the entire DNS system, but rather specific issues within a major service provider’s network:
- Configuration Errors: The most common cause of DNS-related downtime is human error during configuration changes. A single, inadvertent change to a DNS setting can have massive, cascading consequences, as seen with recent outages at AWS and Microsoft Azure.
- Cascading Failures: When a core, foundational service like Amazon’s DynamoDB (which relies on internal DNS for locating resources) experiences a DNS failure, it creates a domino effect. Other dependent services cannot find the resources they need, leading to widespread application and service failures across the internet.
- Caching and Propagation Delays: DNS information is cached (stored temporarily) at various levels to improve speed. Suppose an incorrect DNS record is cached due to an error. In that case, it can take time for the corrected information to propagate (update) across all servers globally (sometimes up to 24 hours), prolonging the outage.
- Systemic Vulnerability: The fact that a seemingly minor, behind-the-scenes system modification can disrupt major global platforms highlights a key vulnerability in critical internet infrastructure.
The Business Impact
The consequences were stark. Banks couldn’t process transactions. Retailers lost sales. Government departments scrambled to restore access to critical systems. For many organisations, the outages meant more than just downtime—they meant lost revenue, reputational damage, and a wake-up call about the risks of cloud dependency.
In the UK alone, services like HMRC’s Government Gateway, Lloyds Bank, and BT were affected. Globally, platforms such as Reddit, Slack, and Snapchat went dark. The economic cost is still being calculated, but early estimates suggest losses in the hundreds of millions, maybe more.
A CISO’s Perspective: Resilience Starts with Architecture
Our own Chief Information Security Officer (CISO), Geraint Williams, summed it up well: “To use a service across the Internet, you must be able to find it—hence DNS. However, like on-premise services, you can only use cloud services if they are available. AWS and Azure recently showed the problem of relying on the cloud for potentially critical services. The alternative for critical services is expensive, internally hosted infrastructure.”
This insight points to a deeper truth. DNS, the protocol that helps users find services online, was designed to be distributed. But cost-cutting and centralisation have eroded that resilience. When DNS fails, everything fails. And when cloud services go down, organisations are reminded that availability is not guaranteed.
What Organisations Can Do to Improve Resilience
The answer isn’t to abandon the cloud. It’s time to rethink how we use it. Resilience must be designed into every layer of infrastructure—from DNS and authentication to data storage and application delivery.
Organisations should consider hybrid or multi-cloud strategies that reduce reliance on a single provider or region. While this adds complexity, it also provides failover options when things go wrong. Distributed DNS services, edge computing, and selective on-premise hosting for critical workloads can also enhance resilience.
But architecture alone isn’t enough. Businesses must regularly test disaster recovery plans, simulate outages, and understand the dependencies within their systems. Resilience isn’t just about uptime—it’s about recovery, continuity, and control.
Resilience Is a Strategic Imperative
Recent outages weren’t anomalies; they were a wake-up call. As digital infrastructure becomes increasingly critical to business continuity, resilience must evolve from a technical concern to a strategic priority. It’s no longer just an IT issue; it’s a boardroom imperative. Organisations must ask: If our cloud provider fails, are we ready?
Modern Networks is more than a service provider; we’re your strategic partner in resilience. We help organisations plan for every eventuality, ensuring that infrastructure is not only smart, secure, and scalable, but also built to withstand disruption. Through robust technical design, proactive planning, and operational preparedness, we can enable our clients to maintain continuity, build trust, and achieve better outcomes—even when the unexpected happens.