By asahifile 
Google Cloud said it restored services at its London region on Tuesday evening after a “power event” at a third-party facility knocked out a cluster of services used by British banks, delivery firms, the court service and the NHS app. The disruption began just after 07:00 local time, when Google’s status dashboard reported “elevated errors and intermittent network failures” in the europe-west2 region, and it took most of the working day to clear.
Revolut, Lloyds Banking Group (Lloyds, Halifax and Bank of Scotland), HSBC and Barclays were among the institutions whose customers reported problems, while Royal Mail, parcel firm DPD, e-bike service Forest and the video platform used for remote court hearings all flagged issues. Google later narrowed the origin to a “power event” inside europe-west2-b, one of the smaller zones inside its London data-centre complex, and by mid-afternoon said the incident had been “mitigated for most of the customers” affected.
What actually failed
Regional cloud outages are usually software failures: a bad configuration pushed to a load balancer, an authentication service timing out, a routing table going stale. This one was different. A power event in a single availability zone is physical infrastructure failing, and it produced a cascade that reached well beyond Google’s own customers.
Revolut said several services were partially unavailable, including card payments, topping up balances by bank card, ATM withdrawals and currency exchange, and stressed that “all our customers’ money and data are safe”. More than 1,500 user reports about Revolut were logged with outage trackers shortly after 08:00. Lloyds said its app and online logins were working normally but that some customers were “having issues with certain online services”; HSBC apologised and noted the problem was “also affecting other organisations”.
One customer, 25-year-old Rachael Ellen from Glasgow, told the BBC she could not pay for parking at a shopping centre and could not see an incoming transfer from a relative. It is a small anecdote with a large point: the failure mode of concentrated cloud infrastructure is not a blank screen, it is a half-working system that quietly refuses individual transactions.
Why one zone mattered so much
Cloud providers deliberately design regions so that a single zone failure degrades rather than destroys a service. Customers are expected to spread workloads across zones, and the largest enterprises do. The outage exposed how many financial and logistics services either run in a single zone or depend on a supplier that does.
“Google Cloud’s outage is a reminder of how deeply our economies depend on just a handful of cloud providers,” said Prof Feng Li of Bayes Business School at City St George’s, University of London. “A problem in one region can interrupt banking, payments and other services well beyond the provider itself.”
That second sentence is the part regulators keep circling. Britain’s banks have spent a decade moving core services to AWS, Azure and Google Cloud, and supervisors have pushed for exit plans and multi-cloud resilience. Yet a six-hour degradation in one London zone still reached payment cards, parcel tracking and court hearings. Separately, ING Germany reported login problems on the same day, though it has not been established that the two incidents were connected.
The resilience bill nobody wants to pay
Redundancy across zones and providers is technically straightforward and commercially unattractive. Duplicating data and compute across two clouds roughly doubles the infrastructure cost of a workload, and it requires engineering teams to accept slower releases and more complex data plumbing. Most boards approve the cheaper design and buy insurance instead.
Google said it would publish a full analysis of the incident on its service health dashboard, which is where the interesting detail will land: whether the power event was a utility failure, a switching fault or a generator problem, and why the blast radius crossed zone boundaries that are supposed to contain it.
What to watch
- Google’s post-incident report on the europe-west2 power event, due on its public status dashboard.
- Whether the UK Financial Conduct Authority or the Bank of England asks affected banks for formal incident reports, and whether any of them disclose how much of their stack sits in a single zone.
- Whether Revolut, Lloyds or HSBC publish customer-remediation commitments, such as waived fees for failed transactions.
- Whether ING Germany confirms or rules out a link to the London outage.
- The next quarterly cloud-reliability disclosures from the three largest providers: outages are becoming a competitive argument, not just an operational one.
Sources: BBC News, CNN/Reuters, ABC News, UPI via Yahoo News.