AWS outage: The event that causes the UK to hit the cloud repatriation button?
Last week’s global outage brought the UK’s reliance on US tech firms back to the fore. CRN asks seven channel partners if customers are asking for alternatives as a result
Monday, October 20 saw many of the world’s biggest apps and websites suddenly grind to a halt thanks to problems at an AWS datacentre in North Virginia.
The hyperscaler had resolved the issues by the next day, but not too soon before the event had reignited the debate around whether the UK is at the mercy of too few US providers.
Cloud repatriation is a topic CRN and the channel have been keeping a watchful eye on.
In August several partner firms gave insights on how curious UK businesses are about repatriation, why it’s a growing discussion, and whether the US has something to do with it.
Following this latest example of the UK falling peril to a US fault, CRN asks Computacenter, Insight UK, Bell Integration, SoftwareOne, Bytes, Axians UK and Krome Technologies whether more customers are asking for alternatives, and their thoughts on whether this is a signal to move away from US hyperscalers.
Paul Nearn, chief technologist, cloud at Computacenter
Since the outage, have you had concerned customers now asking for alternatives? Are the sceptics now convinced, switching from planning to acting?
Over the last five years, AWS and other public cloud platforms have had remarkable levels of uptime, which has driven the substantial use of them by organisations globally.
The recent outage doesn’t change the fact that cloud remains the most resilient way of running infrastructure services available in the market today but it did underscore not only how deeply businesses rely on their digital infrastructure estates but also how interconnected and dependent applications and services are across multiple points of failure.
We talk to our customers about the need for operational resilience to prevent or quickly recover from the likes of the recent cloud outages (and increasing threats like cyberattacks) to ensure they have a clear and concise approach to resilience and are building highly available architectures across cloud and hybrid platforms.
Do you think the AWS outage will and should act as a catalyst for the UK to diversify cloud computing, and move away from the US hyperscalers?
The harsh reality is that outages can occur across any platform, be that in the public cloud, the private cloud or on-premise and should be expected and planned for.
Those who use events like this to support a kneejerk reaction to diversification or repatriation not only risk adding further layers of complexity to their systems but may not understand the time and cost of change such an approach would require.
We see events like these not as a technology driven catalyst for change but as a warning that resilience needs to be top of mind for customers and should be business led, inclusive of security, compliance and connectivity , thus driving the imperative for organisations to build recovery into their business plans.
Do you think channel partners should be leading this charge?
We are helping our customers drive this operational resilience discussion, from fully understanding what their assets are, what dependencies they have on cloud and SaaS providers and how they operate within a multi-platform and hybrid infrastructure model.
We work with customers engineering teams to understand for example to assess if they are correctly using SRE and using disciplines like chaos engineering to find weaknesses and critical dependencies before issues occur thus ensuring fault tolerance and maximum uptime.
Are they successfully using multi-regional or hybrid approaches to failover and protect their data? A partner adds real value by understanding the complexity and the risks and working closely to help a customer both plan for failure and successfully recover from it.
Darren Hedley, managing director at Insight UK
Since the outage, have you had concerned customers now asking for alternatives? Are the sceptics now convinced, switching from planning to acting?
Not as a direct result of the outage, no. We haven't seen so much of that.
We've seen a lot of it through the year, where clients are thinking much more about data residency and where that data sits.
If you think about the AI stack that comes behind it, ultimately, where you get to is, what data have you got and where? How do you need to access it? What's the cleanliness of that data? Is the data good enough to make AI investments?
A recent McKinsey Report stated datacentre spend is going to be one per cent of global GDP going forward.
So that suggests that actually what we're seeing move around data residency, people thinking much more about having that data local.
Does that mean clients are actively pulling data back out of AWS? We're not seeing that. What we're seeing is much more planning around what is fit for purpose going forward.
We called this hybrid cloud a few years ago before we rebranded it into AI. But I think that conversation is still really relevant.
Particularly in EMEA and where we have clients who operate in the UK, France and Germany, there are very different attitudes towards data residency and how that works, particularly around what we've seen with those large hyperscalers.
We haven't seen clients saying they think AWS is now unreliable because of the outage. But we are seeing far more deliberate planning around what needs to be hosted locally, whether that's controlled by the client directly, whether that's from a hyperscaler with residency in the UK.
Lee Alderman, director of service development - cloud, infrastructure & security at Bell Integration
Since the outage, have you had concerned customers now asking for alternatives? Are the sceptics now convinced, switching from planning to acting?
We’re not seeing any evidence of knee-jerk reactions.
From what we saw, the incident was largely treated as an operational outage rather than an event that necessitated invocation of business continuity plans and associated disaster recovery procedures.
Any planning and action plans should be focused on resilient architectures – designing for failure – starting with refreshing your business impact analysis (BIA) and prioritising the systems that represent your minimum-viable-company.
Do you think the AWS outage will and should act as a catalyst for the UK to diversify cloud computing, and move away from the US hyperscalers?
Not as a direct consequence of this outage, no.
Your hosting strategy shouldn’t be driven by a single event that lasted less than a day, there is certainly no need to reactively pivot to repatriation or geopatriation strategies.
It should be a catalyst to review the business resilience that your IT architectures provide. Outages will continue to occur, the hyperscalers have, and will continue to ensure containment.
Cloud adoption and modernisation continue to grow, so too, the headlines when there is an issue – but don’t let this diminish your trust in cloud computing, focus on what an impact would mean to you and design for failure.
Gudmundur Adalsteinsson, chief partner & sales officer at SoftwareOne
Since the outage, have you had concerned customers now asking for alternatives? Are the sceptics now convinced, switching from planning to acting?
In talking to our partners and customers, we see that conversations around cloud strategy, risk management, operational resilience, and data portability are increasingly important. We are committed to supporting organisations in making informed decisions tailored to their unique needs.
Do you think the AWS outage will and should act as a catalyst for the UK to diversify cloud computing, and move away from the US hyperscalers?
Hyperscaler datacentres, both in the US and Europe, offer some of the most secure and resilient infrastructures available. Many governments and organisations rely on them for critical workloads and the platforms are also the most mature for leveraging AI.
Instead of moving away from hyperscalers, there should instead be a nuanced conversation about smart diversification, such as multi-region, multi-cloud, and well-architected solutions that align with specific business needs.
Everybody is going to experience some sort of outage at some point in time, nothing is 100 per cent perfect. It’s how you recover that matters.
At SoftwareOne, we encourage businesses to adopt digital sovereignty, which ensures compliance with local regulations, builds operational resilience, and gives organisations full control over their data and digital infrastructure to protect privacy and security.
Do you think channel partners should be leading this charge?
Channel partners play a critical role in helping organisations navigate the complexity of modern cloud environments. They have a unique opportunity to set the benchmark for building resilient business models that can adapt to the evolving dynamics of the cloud.
Russell Crampin, managing director at Axians UK
Since the outage, have you had concerned customers asking for alternatives? Are sceptics now convinced, moving from planning to acting?
The outage did not affect us directly, so we did not receive any customer complaints.
However, we are actively engaging with several customers on business continuity planning, particularly around cybersecurity risks and supply chain resilience.
It’s no longer a question of if but when an organisation will experience disruption—whether due to a cyber incident or a traditional outage. Having a robust, business-owned continuity plan that can be activated in an emergency is critical to maintaining operations, even at a reduced level.
Do you think the AWS outage will and should act as a catalyst for the UK to diversify cloud computing and move away from US hyperscalers?
A diversified supply chain is essential—not only for resilience but also for commercial flexibility. Data sovereignty is another key consideration when assessing cloud strategies and supplier dependencies.
Do you think channel partners should be leading this charge?
This is fundamentally a business issue. Our role as partners is to educate, advise, and present viable alternatives.
Simon Dobbs, divisional director ‑ cloud sales at Bytes
Since the outage, have you had concerned customers now asking for alternatives? Are the sceptics now convinced, switching from planning to acting?
Whether it's an outage like this latest one or a cybersecurity incident, we are increasingly seeing our customers recognise that they need a credible plan to ensure services remain available and can be recovered quickly and effectively. This year has shown just how susceptible even the biggest players are.
Building highly resilient and redundant systems is expensive and it has historically been difficult for customers to get sign off when their board sees the cost of building that resiliency. We are now seeing this change as Execs begin to understand how vulnerable they are to a catastrophic outage.
Relying on someone else isn't a viable strategy.
Do you think the AWS outage will and should act as a catalyst for the UK to diversify cloud computing, and move away from the US hyperscalers?
Given that the majority of infrastructure vendors are also American, I think we need to be pragmatic about how achievable this is in reality. The servers could reside in the UK, but they would still be built and the code developed elsewhere.
What it does underline to us and something we advocate to our customers is ensuring that they have a clear and solid understanding of their critical infrastructure. What is it, where is it and how do you recover it?
And what happens if there is an outage that wipes out your primary site, be that a hyperscaler, your hosting provider or your own datacentre. Having a strategy that spans a couple of those targets does give you options.
Do you think channel partners should be leading this charge?
Our job always has always been and always will be to help customers identify and build the services most suited to their organisation.
Rupert Mills, CEO at Krome Technologies
Since the outage, have you had concerned customers now asking for alternatives? Are the sceptics now convinced, switching from planning to acting?
We have had one or two clients asking more about how to get a more effective hybrid. On an unrelated note the migration of our clients from VMWare to HyperV (since the Broadcom acquisition) has made this hybrid approach more easily achievable and gives clients the ‘best of both.’
Do you think the AWS outage will and should act as a catalyst for the UK to diversify cloud computing, and move away from the US hyperscalers?
Ultimately, I suspect the UK will struggle to mount a credible challenge to the US hyperscalers with the scale of the market here and the business environment for anyone wanting to invest in it.
It would be good to have offerings available and I'd like to see it, but datacentres seem to be more likely to lease their space from the US corporates in the event they have the capacity.
With demand for power from AI, this is even more of a challenge now.
Do you think channel partners should be leading this charge?
Channel partners need a service offering to take to our clients, we do this with HyperV and Azure hybrid for example, but creating something viable in terms of a shared cloud service is tricky at scale so would need someone with deep pockets to get it off the ground.