A power problem caused an outage in Google Cloud’s London infrastructure on Tuesday, October 20, affecting several computing, storage and networking services. Google subsequently announced a gradual recovery of services. On the same day, Revolut reported partial unavailability affecting functions including card payments, transfers and access to accounts.
The Google Cloud incident began at 09:08 Romanian time, according to the official timeline. The company located the outage in the europe-west2-b zone, part of its London region, named europe-west2. Initial messages described elevated error rates and intermittent network failures, before subsequent updates identified the power problem behind the infrastructure outage.
Power was restored before services recovered
In the first message in the log, published at 11:35 Romanian time, Google said engineers were investigating the incident and that no workaround was available for customers at that point. Updates at 12:24 and 14:00 continued to describe the investigation as ongoing. During this stage, the company added moving resources to another zone or region among the options for customers able to make that change.
At 14:12 Romanian time, Google stated that the power problem had been mitigated and that infrastructure services would recover next. In its message published at 14:58, the company said it was prioritizing existing workloads. Creating new resources could remain unavailable at this stage, even as other functions began to return to service.
The update published at 15:44 recorded that services for most workloads in europe-west2 were recovering as of 15:10. These times are expressed in Romanian local time. Google’s log uses US Pacific time, placing the start of the incident on the evening of October 19 there, although it was already the morning of October 20 in Romania.
In its description of symptoms, Google distinguished unavailable services in the europe-west2-b zone from elevated errors and intermittent network issues that could occur across the europe-west2 region. The company retained this description in the 15:44 update, alongside the information about service recovery for most workloads running in the affected region.
The notice covers infrastructure, databases and applications
The incident’s product list includes Google Compute Engine and Google Kubernetes Engine, services used to run applications, alongside Cloud Run and Google App Engine. Database and storage components are also listed, including Cloud SQL, Cloud Spanner, Cloud Firestore, Cloud Storage and Cloud Filestore.
The same list includes analytics and data integration services such as BigQuery, Dataflow and Datastream, together with Pub/Sub, Cloud Load Balancing, Cloud NAT and Virtual Private Cloud. In its updates, Google explicitly stated that Looker Studio was unaffected. Looker for Google Cloud appears separately among the products included in the incident notification.
Within Google Cloud’s structure, a region is a geographic area consisting of several zones. A zone is a deployment location for resources within that region. The company’s documentation distinguishes zonal resources, such as certain virtual machines and disks, from regional or global resources. In Tuesday’s incident, europe-west2-b identifies the London zone involved in the outage.
For affected customers, Google suggested moving workloads to another available location, either a zone or a region, where that operation was feasible. The message concerns relocating the technical components running on affected infrastructure. It is separate from the stages in which Google restores power and recovers existing services within the affected location.
Revolut also reported issues affecting business accounts
Revolut’s system-status page for Romania displayed a partial-unavailability indicator for 31 components when checked at 16:00 Romanian time. Listed functions included sign-up, sign-in, card payments, card-based account top-ups, transfers and ATM withdrawals. The indicator describes platform components, while the published status label for the incident identifies the availability problem as partial.
The list included currency exchange, stock trading, cryptocurrency operations and some card-related functions, such as ordering a card and viewing its details. Identity verification, credit services, live support chat and the Revolut website were also listed on the status page. The bank’s information therefore covered several categories of operations as well as access to support.
For businesses, the same page listed Revolut Business registration and sign-in, outgoing transfers, the mobile app and the web app. It also included the Business API, merchant API payments and the Sandbox testing environment. The disruption reported by the bank therefore concerned both interfaces used directly and integration services involved in companies’ day-to-day operations.
Revolut stated that its teams were working to restore functionality and that customers’ funds and data were safe. The bank said work had begun after the problem was identified during the morning and apologized to affected customers. On its status page, services for personal customers and businesses are displayed together within the same service overview.
Google Cloud is a Revolut infrastructure partner
The companies’ technology relationship is documented in an expanded partnership announced by Revolut on September 4, 2025. The bank explained then that it used Google Cloud infrastructure to develop its global systems, from backend operations to the experience in its customer-facing app. The announcement described a partnership extending over multiple years.
The same announcement mentioned using Google’s artificial intelligence and machine-learning tools, including Gemini models, to improve fraud detection and develop more relevant product offerings for customers. The expanded partnership covered both the infrastructure running Revolut’s systems and the use of these tools in the development of the bank’s services.
At 16:00 Romanian time, the latest update in Google’s log was the one published at 15:44. While Google described service recovery for most workloads, Revolut’s status page for Romania continued to display partial unavailability. Google had announced that it would communicate further details about the incident by 17:00 Romanian time.
