A recent power outage at a Google Cloud datacentre in London significantly impacted various UK services. The disruption quickly escalated, affecting several crucial sectors. Consequently, multiple banking operations faltered, court hearings experienced delays, and patients reported temporary issues with the NHS application.
This particular datacentre is located within Google’s europe-west2-b zone. This is one of over a dozen such “zones” maintained by the US technology giant across Europe. These zones serve its extensive client base.
The outages, which according to reports, impacted Royal Mail, Revolut, HSBC, and Lloyds Bank, among others, were not as severe as a similar incident last year. That prior cascade saw Amazon Web Services go offline for nearly half a day.
That earlier outage affected a much larger node, Amazon’s us-east-1, situated in a highly concentrated region of Virginia. This area hosts the highest density of datacentres globally. Its failure caused ripple effects worldwide, taking out the Roblox game and the Signal messaging app. It even caused smart mattresses globally to overheat overnight.
Today’s problems largely peaked around 9 am. Reports suggest that services were mostly resolved by mid-afternoon. Since Google Cloud manages data and operations for numerous businesses, the knock-on effects were extensive, though primarily confined to UK businesses.
Revolut confirmed “a service disruption affecting multiple services.” However, the company assured customers that “all our customers’ money and data are safe.” Lloyds stated its app was functional. Nevertheless, it acknowledged that “some [people] are having issues with certain online services this morning.” HM Courts and Tribunal Service, which handles remote video hearings, also confirmed an issue with its Cloud Video Platform. It actively worked with partners to restore service.
These events underscore a growing concern about deep UK cloud reliance on a few key technology firms. Internet infrastructure expert Michał Woźniak highlighted the increasing concentration of vital services on the infrastructure of these “hyperscalers.”
Woźniak remarked that hyperscalers have become “too big to not fail.” Moreover, he noted their deep integration into our digital lives, making any switch seem unimaginable. This situation exposes sensitive information to various risks, including outages.
Woźniak further commented on the repeated failures. He argued that it is unreasonable for a handful of giant companies to cause public institutions or vast parts of the digital economy to cease functioning en masse. He also pointed to Google’s increased use of AI, with 75% of new code now AI-generated. This raises questions about reliability, especially after previous AWS outages were linked to internal AI tools.
A Google spokesperson confirmed the resolution of a service disruption. This disruption occurred at a third-party facility in its europe-west2 cloud region due to a power event. Services have been restored, and an analysis of the incident will follow.
