A power failure inside one of Google Cloud’s London data centre zones knocked dozens of cloud services offline for hours on Tuesday, disrupting UK banks, the postal service and public-facing websites before Google restored full service later in the day. The outage began at 06:08 UTC in europe-west2-b, one of three availability zones in Google’s London region, and hit 38 Google Cloud products, according to Google’s status page and reporting on the incident.
Among the customers caught up in the failure were Lloyds Bank and HSBC, which both confirmed their customers had trouble reaching banking services, and the digital bank Revolut, where ATM withdrawals, bank transfers and card top-ups were disrupted. Royal Mail, delivery firm DPD, property site Rightmove, BT and Screwfix were also affected, and Downdetector logged hundreds of problem reports from UK businesses through the morning.
What Google Says Happened
Google blamed the outage on a “power event” inside the zone. “A power event caused an outage within europe-west2-b zone. The power event is mitigated. We are seeing the recovery of services and are prioritizing the recovery of existing workloads,” the company said on its status page, warning that customers might temporarily be unable to create new workloads. Services that went down included BigQuery Data Transfer Service, Cloud Filestore, Cloud Run, Google App Engine, Google Cloud NetApp Volumes, Google Kubernetes Engine and Managed Service for Apache Kafka.
The recovery came in stages. Google said most services had returned to normal by 12:10 GMT, and its status page said the issue was mitigated for most customers by 14:03 UTC, with a very small number still affected while engineers finished the work. A later wire report confirmed full availability had been restored across the affected services. Google said it will publish an analysis of the incident once its internal investigation is complete.
Banks Confirm The Customer Impact
Lloyds told customers on Tuesday afternoon that it understood the root issues had been resolved and expected all its services to be back to normal soon. HSBC said its customers had experienced “difficulties accessing our services” and that it would monitor a situation also affecting other organisations. A Revolut spokesperson apologised for the disruption, said the bank had been working to fix the issue since it was identified, and confirmed that all customers’ money and data were safe.
The customer list is the story’s sharp edge: banks, the national postal operator and a major property portal all depend on the same London cloud region, so a single zone’s power problem became a consumer-facing failure across unrelated industries within the same morning.
One Zone, Thirty-Eight Products
The technical shape of the failure is familiar to cloud engineers. An availability zone is meant to be an isolated failure domain inside a region, but workloads that live in one zone — or that depend on regional services with zonal components — still go down with it. Dr Amro Al-Said Ahmad, a lecturer in computer science at Keele University, told The Register that a core region failure propagates through applications, APIs, databases, authentication systems and third-party software several layers removed from the original fault, and that resilience requires designing for regional failure rather than assuming the cloud removes it. Google’s own hardware work this week pointed the other way — toward more capable infrastructure — as the company completed its open-source Coral NPU design for edge devices.
Why It Matters
The London outage matters because it shows how concentrated the UK’s digital plumbing has become. Google Cloud is now significant enough to British banking and public services that UK regulators placed it under direct oversight as a critical third-party supplier earlier this year — and a power event in one zone still reached high-street banks, a challenger bank and the postal service at the same time. For the businesses affected, the incident is a live test of the multi-zone and multi-region designs that are supposed to contain exactly this kind of failure. For Google, which is competing for regulated-industry workloads against Microsoft and Amazon, the promised post-incident analysis will be read closely: the outage follows other recent scrutiny of its AI services, including rulings in Germany holding Google liable for false claims in its AI answers, and it lands while the company is pushing its AI search products deeper into everyday use.
Background
Google Cloud’s London region, europe-west2, is one of the main homes of UK financial-services computing, and cloud concentration risk has been a standing concern for British regulators as banks have moved core systems onto a small number of US providers. Outages at major clouds have repeatedly shown the same pattern — a localised infrastructure fault cascading into national consumer disruption — which is why Google Cloud’s designation as a critical third-party supplier, and the post-incident report it has now promised, carry weight beyond one morning’s downtime.
Source: The Register; Dawn (AFP); TradingView News (dpa-AFX)
0 Comments