Google Cloud London Power Cut: How One Zone Paralyzed UK Banks, Courts and Couriers

What happened

At around 7:05 BST on Oct 20, Google Cloud's europe-west2-b availability zone in London reported a power cut, and multiple cloud services soon showed elevated errors and intermittent network failures. Citing Google's engineering updates, Computer Weekly reported that backup generators on site failed to engage even after mains power returned, forcing engineers to keep draining frontends and backends from the zone's clusters while onsite technicians verified the power systems.

The disruption peaked around 9:00 a.m.; Google said the power event was mitigated at 12:27 and most workloads had recovered by 14:48, closing the incident on its status page at about 15:00 BST, for a total of roughly 7 hours 52 minutes. The Ministry of Justice confirmed at 15:52 that the courts' cloud video platform had still not fully recovered.

Google described the event as a power incident at a third-party facility in London, said services had been restored, and promised an analysis after its internal investigation. The same region suffered a cooling failure during the UK's record heatwave in July 2022, when multiple redundant cooling systems failed at once and engineers powered down part of the zone to save the hardware, terminating roughly a third of its virtual machines. Unlike four years ago, europe-west2 is now Google's only cloud region in the UK, leaving customers with no second UK region to fail over to.

The blast radius

Banks and fintech took the heaviest hit. Revolut marked 27 services as partially unavailable or down — logins, card payments, ATM withdrawals, transfers, currency exchange, stock and crypto trading all halted — covering its 13 million UK customers (plus 3.4 million in Ireland); Downdetector logged thousands of user reports against Lloyds, Halifax, HSBC, Barclays and Bank of Scotland, while Computer Weekly's compiled list also named NatWest, Nationwide, budgeting app Emma and credit-reference firm Equifax.

Public services went down at the same time. The cloud video platform used by courts in England and Wales failed across large parts of the country from 10:00 a.m., and the Criminal Bar Association warned that trials and hearings "can not proceed"; a judge at Southwark Crown Court mentioned "problems with Google" before the platform recovered. NHS England and gov.uk services appeared on the outage lists, alongside Royal Mail, DPD, Rightmove, Screwfix and BT.

Computer Weekly called it another reminder that the UK's largest businesses remain highly exposed to cloud outages: because europe-west2 is Google's only UK region, customers without multi-region resilience had nowhere to move within the country. One engineer wrote on Reddit that the peer zones a and c "didn't pick up the slack completely," for reasons unclear. Computer Weekly also noted a similar episode last autumn, when an AWS outage disrupted UK tax-office and banking services at once.

What each side said

A Google spokesperson told Reuters: "We have resolved a disruption to Google Cloud services at a third-party facility in our europe-west2 cloud region due to a power event. Services have been restored. We will update our public status dashboard with an analysis." Google also advised affected customers to "try to migrate their impacted workloads to another available location (zone or region)" if feasible.

Revolut said on X that day: "We are currently experiencing a service disruption affecting multiple services. We can confirm that all our customers' money and data are safe. We've been working to fix the issue since it was identified this morning and apologise to anyone affected." The company separately told media it was reviewing the incident to reduce the impact of future third-party disruptions, and that customer funds and data stayed safe throughout.

A European Banking Authority spokesperson told Reuters it was in contact with Google Cloud to "obtain further information on the incident, understand its implications and assess its potential impact on financial entities in the EU." Reuters described the outage as the first major test of new European rules designed to strengthen the resilience of technology providers, including cloud firms, on which the financial sector depends. The UK Prime Minister's office confirmed it was in contact with Google; the Financial Conduct Authority and the Prudential Regulation Authority said they were engaging with Google and affected firms, stressing customers must not be left out of pocket; House of Commons Treasury Committee chair Meg Hillier demanded details of the regulators' response and the financial impact by 4 November.

Sam Richardson, editor of Which? Money, said people would be "rightly concerned" about outages that could "prevent people making essential purchases or even paying their bills," adding that "tech and banking outages have become increasingly frequent in recent years, and the impact can be widespread when they happen."

Key facts

1. Where it started: At around 7:05 BST on Oct 20, the europe-west2-b availability zone lost power; citing Google engineering updates, Computer Weekly reported that backup generators on site failed to engage. Google attributed the event to a power incident at a third-party facility in London.
2. Duration: Google said the power event was mitigated at 12:27 and most workloads recovered by 14:48; the status page closed the incident at about 15:00 BST, for roughly 7 hours 52 minutes. The Ministry of Justice said at 15:52 that the courts' cloud video platform had still not fully recovered.
3. Revolut: 27 services went down, affecting 13 million UK customers; the company stressed funds and data were safe and said it was reviewing the incident to reduce the impact of future third-party failures.
4. Other banks and firms: Downdetector logged user reports against Lloyds, Halifax, HSBC, Barclays and Bank of Scotland; Computer Weekly's compiled list also named NatWest, Nationwide, Emma and Equifax. Royal Mail, DPD, Rightmove, Screwfix, BT, NHS England and gov.uk all appeared on outage lists.
5. Judiciary: The cloud video platform used by courts in England and Wales failed from 10:00 a.m.; the Criminal Bar Association said trials and hearings "can not proceed"; a judge at Southwark Crown Court mentioned "problems with Google" before the platform recovered.
6. Regulators step in: The European Banking Authority contacted Google over the incident; the Prime Minister's office stayed in touch with Google; the FCA and PRA engaged with Google and affected firms; the Treasury Committee demanded a report by 4 November.
7. Structural backdrop: europe-west2 is Google's only UK cloud region; peer zones a and c did not fully absorb zone b's load. In July 2022 the same region lost power after multiple redundant cooling systems failed at once.