Skip to content
When the Cloud Goes Dark: Why On-Premises Storage Still Matters
by Gavin Sutton, Head of Marketing
Yesterday, one of the largest cloud infrastructure providers experienced a major operational failure that rippled across dozens of high-profile websites, apps, banking platforms, and public services. From mobile banking to entertainment platforms, many organisations found themselves offline through no fault of their own. At the same time, those with alternative infrastructure approaches were better positioned to maintain continuity.
This outage shouldn’t just be read as an isolated failure of cloud services, it serves as a real-world reminder that cloud alone is not always enough. For many businesses, especially those in regulated industries like defence, financial services and media, the question isn’t “Should we move everything to the cloud?” but rather “How do we build infrastructure that keeps us online, secure, and resilient when the unexpected happens?”
The Cloud Outage: What Happened
The incident largely traced back to the AWS US-EAST-1 region and later affected global traffic and services. Banks such as Lloyds Bank, Halifax and Bank of Scotland reported service disruption, while apps like Fortnite, Snapchat, Canva and smart-home services also went offline. The root causes cited increased error rates and latencies in core cloud services and the incident underscores the fundamental reliance on large hyperscale cloud providers.
Why On-Premises Storage Isn’t Dead
When the cloud fails, a complementary or alternative on-premises infrastructure can make the difference between business continuity and a full outage. Key advantages include:
Localised Control & Independence
On-premises infrastructure means that customers retain full control over their hardware stack, data flows and failure domains. They’re not at the mercy of the “shared fate” of other customers in a multi-tenant cloud region. In a world where the cloud provider’s region goes down, local control matters.
Predictable Performance & Latency
Cloud regions suffer from shared infrastructure, ‘noisy neighbours’ and network dependencies. When latency and predictability matter such as in financial services, post-production, or defence systems, on-premises or edge-located storage and compute deliver consistent results.
Sovereignty, Compliance & Data Residency
For clients with regulated or national-security concerns, keeping data and systems onsite or within a sovereign region is essential. A hyperscaler outage can trigger cascading issues across clients and services in its ecosystem, with on-premises you sidestep many of those shared-risk pitfalls.
Lower Risk of External Failure Cascades
A single outage in a cloud data-centre can impact thousands of customers. If businesses operate their own infrastructure or use a well-architected private or hybrid cloud, they’re not sharing that risk. The AWS outage is a textbook example of how broad the impact of a hyperscaler failure can be.
Hybrid Cloud & On-Prem: The Balanced Approach
Of course, we’re not saying the cloud has no place, far from it. The smart strategy today is hybrid – leveraging cloud for elasticity, global reach and specialised workloads while relying on on-premises or edge infrastructure for core systems, guaranteed performance, and resilience.
Here’s how a hybrid model might look:
- Core applications and sensitive data processed on-site
- Scaling demanding workloads to cloud when required
- Critical data mirrored across on-premises and cloud for continuity
- Edge compute and storage placed near where data is generated
In doing so, organisations avoid putting “all their eggs in one basket”.
What This Means for Storage & Infrastructure Decision-Makers
- Evaluate whether critical workloads can continue to function when a major cloud provider fails.
- Ask: If the cloud region fails, what’s the fail-over path?
- Ensure storage architecture delivers low latency, high throughput and local performance SLAs.
- Consider leveraging high-density on-premises storage, especially for large datasets and archives, as a modern strategy.
- Build a hybrid model with sovereignty, local control and cloud agility in mind.
Final Thoughts
The recent cloud outage is more than just an inconvenience, it’s a wake-up call for IT decision-makers who believed the cloud was infallible. When banks, apps and services go offline because one cloud provider faltered, the value of resilient, localised infrastructure becomes crystal clear.
If your customers are looking to build AI-ready, performance-driven, resilient infrastructure, it’s time to think beyond “cloud only”. By combining high-performance on-premises storage, smart edge deployments and thoughtful cloud integration, they can create a future-ready architecture that keeps their business running no matter what happens upstream.
Do your customers need help assessing their infrastructure strategy and how to deepen on-premises or hybrid resilience? Contact us to explore how we can support them.