The Google Cloud London outage on 20 October 2026 knocked out banking apps, court video links and parcel tracking, all because of a single power failure in one data-centre zone. Google says the power event is fixed, and it has promised a full analysis. If you run anything online, the lesson is simple: one zone is not a backup plan.
Here is the short version before the detail.
- When: Tuesday 20 October 2026, with Google saying customer impact began around 06:08 UTC.
- Where: the europe-west2-b zone, one of three zones in Google’s London region.
- Who noticed: Revolut, Lloyds, HSBC, Royal Mail, DPD, Rightmove and the courts’ video system, among others.
- Cause: Google calls it “a power event.” No detailed report is out yet.
If you only wanted to know whether your bank app was broken because of this, that is the answer. Keep reading if you want to know what it means for you.
What happened during the Google Cloud London outage
According to The Register, Google’s status page said a power event caused an outage in europe-west2-b. Customer impact began at 06:08 UTC and most services were back by about 14:03 UTC. That makes it a long morning for anyone who needed their app to work.
The timeline below comes from Computer Weekly, converted to UTC.
| Time (UTC) | What happened |
|---|---|
| ~06:00 | Power loss in europe-west2-b |
| 09:57 | Google confirms an issue with multiple services in europe-west2 |
| 10:05 | Revolut tells customers their money and data are safe |
| 11:27 | Google says the power event is mitigated |
| 13:48 | Google reports recovery for most London workloads |
| 14:10 | Google says only a “very small number” of customers are still affected |
Not every outlet agrees on the start. ITV News reported Google saying it began late on Monday evening, while the status-page timing points to Tuesday morning. Treat the start time as approximate until Google publishes its report.
Who felt it, and why you probably noticed
The damage spread far beyond Google’s own customers, because many well-known services quietly run on Google Cloud. Computer Weekly lists banks, couriers, retailers and public services among those hit.
| Sector | Reported as affected |
|---|---|
| Banking and fintech | Revolut, Lloyds, HSBC, Barclays, NatWest, Halifax, Nationwide, Equifax |
| Delivery and retail | Royal Mail, DPD, Screwfix |
| Property and web | Rightmove, BT, SiteGround |
| Public sector | Courts video platform, NHS England, Gov.uk |
Reports varied by service, and some were user complaints rather than company confirmations. The courts’ video platform going down is the one that mattered most: the Criminal Bar Association said trials and hearings could not proceed.
Why one power event hit so many services
A cloud region is split into zones so that one failure should not take everything down. europe-west2 has three. Services built to spread across all three should have survived, and many that were tied to the failed zone did not.
The cause is still unclear. Computer Weekly cites an administrator relaying a Google engineering update that mains power returned but backup generators did not engage. Google’s public statement does not confirm that, so treat it as unverified.
There is a second problem. London is Google’s only UK region, so firms that want data kept in the UK have no second UK location to fail over to. That is a real trade-off between data rules and resilience. For a related look at how fragile cloud-dependent systems can be, see our piece on the ConnectWise cloud backup breach.
What to do if you run a site or app
You do not need to rebuild everything. A few checks this week will tell you how exposed you are.
- Check which zone each of your services runs in, and whether any critical one sits in a single zone.
- Spread databases and load balancers across at least two zones.
- Decide whether you need a second region, even if that means a non-UK one for failover.
- Run a real failover test, not just a plan on paper.
Why this matters: most outages are not exotic attacks. They are ordinary failures, like power, that hit a design with no spare.
The same logic applies to AI-heavy workloads that depend on a few huge sites, as we covered in Nvidia’s DSX POD plans.
What you can do as an ordinary user
Mostly, wait. Revolut told customers their money and data were safe, and outages like this usually affect access rather than balances. Avoid repeated payment attempts, which can create duplicates.
If you were charged twice or a payment looks stuck, check your transaction history first. Then contact your provider through its official app or website, not a link in a message.
Frequently Asked Questions
Was the Google Cloud outage a cyberattack?
Google has not said so. It describes a power event, and UK Power Networks said its London systems ran normally, which rules out a substation fault, per Computer Weekly.
Is Google Cloud still down in London?
No. Google said service had mostly recovered by about 14:03 UTC on 20 October, with a very small number of customers still affected at the time.
Was my money at risk?
Revolut said customers’ money and data were safe. The problems reported were failed payments, sign-ins and app access.
Will Google publish a report?
Yes. Google said it will publish an analysis once its internal investigation is done, but it has not given a date.
Our take: treat one zone as one point of failure
This outage says less about Google than about how most of us build. If your app, bank or business needs to stay up, a second zone is the cheap fix, and a tested second region is the real one. Wait for Google’s report before judging the cause, but do not wait to check your own setup.
Leave a Reply