Quick answer: on 20 October 2026, a failure in one of Google Cloud’s three London data centre zones (europe-west2-b) knocked out or slowed services for hours. Banking apps and websites including Lloyds, HSBC, Revolut, Royal Mail, BT and Rightmove were reported as disrupted. Google linked the problem to a “power event”. Revolut said customer money and data stayed safe.
It’s a reminder that “the cloud” is still physical buildings full of servers, and a useful lesson for any business that runs online.
What happened
- The problem began at about 06:08 UTC (11:38 AM IST) in zone europe-west2-b.
- Google’s status page reported “elevated errors and intermittent network failures” across the London region.
- Affected Google Cloud products included Cloud Run, Google Kubernetes Engine, App Engine, Cloud Filestore, BigQuery Data Transfer, NetApp Volumes and Managed Kafka.
- Outage trackers showed problems at Lloyds Bank, HSBC, Revolut, Royal Mail, DPD, BT, Screwfix and Rightmove.
- Google first said there was no workaround, then advised customers to move workloads to another location where possible.
What is a cloud “region” and “zone”?
Big cloud providers split the world into regions (for example, London or Mumbai). Each region has several zones: separate data centres with their own power and cooling. The idea is that if one zone fails, apps running in other zones keep working.
The catch: that protection only works if your app is actually spread across more than one zone. Many companies run critical systems in a single zone to save cost or for simplicity, and when that zone goes down, so do they.
Why a single zone failure spread so widely
- Single-zone setups: some services ran in only one zone.
- Hidden dependencies: an app spread across zones can still rely on one database, file store or login service that lives in a single zone.
- Shared region services: some cloud services are regional, so a big enough zone problem can ripple across the region.
Lessons for businesses, including small ones
- Know where your site runs. Ask your host or developer which region and zone your website, database and backups are in.
- Spread critical systems across zones, or at least keep a ready-to-start copy in another zone.
- Keep backups in a different region from your main site.
- Use a CDN so static pages and images can still load when your server is struggling.
- Have a status page and a plan for telling customers what’s happening during an outage.
- Test your failover at least once. A backup plan that has never been tried is a guess.
Could this happen in India?
Yes. Google Cloud, AWS and Microsoft Azure all run regions in India (Mumbai, Delhi, Hyderabad, Pune and Chennai, depending on the provider). The same rules apply: the zone design protects you only if your setup uses it. Indian banks and fintechs follow RBI rules on business continuity, but many small businesses and online stores run on a single server with no failover at all.
Frequently asked questions
What caused the Google Cloud London outage?
Google linked it to a “power event” affecting zone europe-west2-b in its London region on 20 October 2026.
Which companies were affected?
Outage trackers reported problems at Lloyds, HSBC, Revolut, Royal Mail, DPD, BT, Screwfix and Rightmove, among others.
Was customer money at risk?
Revolut said customer money and data remained safe. Outages like this affect access, not balances.
How can a business protect itself from cloud outages?
Run critical systems across multiple zones, keep backups in another region, use a CDN, and test your failover plan.
Sources: The Register; TechRadar; ABC News; Google Cloud Service Health; AI Weekly.
