Google Cloud's London Zone Lost Power: What Owners Should Know

On 20 October 2026, a Google Cloud zone in London lost power. The zone is called europe-west2-b. A zone is one data center area inside a cloud region. If your website, checkout, or customer portal lived only in that zone, it stopped when the power did. Customers who needed that site would have found it unavailable.

I do this planning with owners in the US and Europe. We find the zone that carries your orders and we prepare a second place that can pick up the work. The Able Retainer is $799 a month for eight hours of senior production time, planned with you. Email greg@ableadmins.com, or start on the infrastructure services page.

What Google reported

Google published its incident report on 27 October 2026. In that report, Google says the data center hosting europe-west2-b lost primary utility feed A at 22:53 US/Pacific on 19 October 2026 (06:53 BST on 20 October) because of an external underground cable fault. Feed B was already locked out for repairs after issues found in a quarterly generator load test on 14 October 2026.

The generators stayed off. Google says the dual-redundant PLCs did not start them, because their undervoltage sensing inputs were wired upstream of the isolated feed B breaker and still saw voltage. A PLC is the controller meant to notice a power loss and start the generators. The static UPS batteries, rated for 5 minutes, depleted. A static UPS is a short-term battery backup. Hosts, local storage, and network aggregation hardware in europe-west2-b lost power at 23:00 US/Pacific (07:00 BST).

The disruption lasted 6 hours 8 minutes, Google says, from 23:00 US/Pacific on 19 October to 05:08 US/Pacific on 20 October (07:00 to 13:08 BST). By then, Compute Engine virtual machines were 99.5% operational, and creating new virtual machines was fully supported. A virtual machine is a server you rent in the cloud. Google says the impact stayed inside europe-west2.

When the other London zones got busy

Zonal resources such as Compute Engine virtual machines require customers to provision failover or cross-zone recovery, Google says. Failover is a second, ready place for the same work. Cross-zone recovery means you can stand that work up in another zone. That work sits with you.

Some failovers failed, Google says, because of capacity constraints and difficulty replicating storage. Replicating storage is keeping a usable copy of your disks somewhere else. Zones europe-west2-a and europe-west2-c saw capacity exhaustion, control-plane API timeouts, and HTTP 500 and 503 errors from redirected traffic. The control plane is the API you use to create and manage servers. An HTTP 500 or 503 is the service answering that it cannot complete the request.

What the BBC reported that day

On 20 October 2026 the BBC reported a Google Cloud spokesperson saying the company had resolved a disruption at a third-party facility in europe-west2 due to a power event and that services had been restored. Revolut told the BBC some of its services were disrupted by the outage and that customers' money and data stayed safe. Guy Baybutt of The Veg Box Company told the BBC the firm could not access its website to process orders or book parcels on DPD during the outage.

For an owner, the scale is familiar: a website, a payment flow, or a shipping desk waiting on one zone.

What you can review this week

A few plain checks show whether one zone is holding your day together.

In your cloud console, read the zone on the servers that take payment, store customer files, or run the database. If they all name europe-west2-b, that zone is your single home.
Read your failover note against a 07:00 BST clock, and refresh any step that is still untested.
Ask where the disks live. Google says some failovers failed because of capacity constraints and difficulty replicating storage. A new server in another zone serves yesterday's orders when a copy of the disk is already there.
Expect the next zone to be busy. Google says europe-west2-a and europe-west2-c saw capacity exhaustion, control-plane API timeouts, and HTTP 500 and 503 errors from redirected traffic.
Draft one calm sentence for customers. Revolut told the BBC that some services were disrupted and that money and data stayed safe. Write the version that is true for you.
Make sure two trusted people can reach the cloud console, the domain, and the payment account. 07:00 BST is 23:00 US/Pacific, late evening on the West Coast.

If you would like me to walk those checks with you, begin with infrastructure services.

How Able can help

I am the engineer who does this work with you. We look at the zone your site, mail, and files use, and we practice the move to a second zone on a quiet day. Infrastructure services include Linux, Windows, AWS, Kubernetes, and mail. We start with a conversation.

The Able Retainer is $799 a month for eight hours of senior production time, planned with you. Email greg@ableadmins.com or call (832) 977-1356. When you want to see how that conversation starts, visit infrastructure services.

Sources

If your business depends on a single cloud zone, I would be glad to look at it with you. You can reach me at greg@ableadmins.com, and you can start with infrastructure services.