Google Cloud says a 'power event' caused today's outage

Google Cloud has blamed a “power event” for taking down multiple online products for customers in Western Europe, and is pledging to restore normal service within the hour.

The unplanned downtime for customers began at 0608 UTC when europe-west-2-b, one of three local availability zones in the London region, started playing up.

According to an update on its status page, Google Cloud said at 1158 UTC that full recovery workloads should be “completed, tentatively” by 1300 UTC. Customers can expect another update at that point, with Google either giving the thumbs up, thumbs sideways or having to admit engineers are still working on a fix.

Services that went AWOL included BigQuery Data TransferService, Cloud Filestore, Cloud Run, Google App Engine, Google Cloud NetApp Volumes, Google Kubernetes Engine, and Managed Service for Apache Kafka.

So what caused online services to melt at customers including Lloyds Bank, HSBC, Royal Mail, Rightmove, DPD, BT, Cloudflare, Screwfix and Revolut? Power.

"A power event caused an outage within europe-west2-b zone. The power event is mitigated. We are seeing the recovery of services and are prioritizing the recovery of existing workloads.Customers may not be able to yet create new workloads."

The Register has asked Google for further elucidation but it said the only information is available on the status page.

The "fragility" of the cloud model is again on display, Dr Amro Al-Said Ahmad, a lecturer in computer science at Keele University told The Register. He estimates that a total of 38 services were down and out for the count.

"The incident highlights the dependency effect of modern cloud ecosystems. When a core cloud region experiences failures, the impact extends far beyond the directly affected services. Applications, APIs, databases, authentication systems, payment platforms, analytics pipelines, and third-party SaaS products that depend on those services can all experience cascading failures."

"A problem in a single cloud region can therefore propagate through an entire digital supply chain, affecting organisations that are several layers removed from the original outage. While this is a regional outage centered on europe-west2, the breadth of affected services highlights a key cloud engineering lesson: resilience is not achieved by simply moving to the cloud. True resilience requires designing for regional failures through multi-region deployments, workload portability, and tested disaster recovery procedures." ®