paas and iaas
London’s Fells area, taking banks, parcel companies and property sites with it – and recovery is ‘tentatively’ expected by 1300 UTC
Google Cloud has blamed a “power event” for the removal of several online products for customers in Western Europe and is pledging to restore normal service within the hour.
THE Unplanned customer downtime began at 0608 UTC when europe-west-2-b, one of three local availability zones in the London region, came into play.
According to an update of his status pageGoogle Cloud said at 11:58 UTC that full recovery workloads should be “complete, provisionally” by 1:00 PM UTC. Customers can expect another update at this point, with Google either giving it a thumbs up, moving aside, or admitting that engineers are still working on a fix.
Services that have disappeared include 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 has caused online services to melt away at clients such as Lloyds Bank, HSBC, Royal Mail, Rightmove, DPD, BT, Cloudflare, Screwfix and Revolut? Power.
“An electrical event caused an outage in the europe-west2-b zone. The electrical event is mitigated. We are witnessing the resumption of services and prioritizing the recovery of existing workloads. Customers may not yet be able to create new workloads.”
The register asked Google for further clarification, but said the only information was available on the status page.
The “fragility” of the cloud model is visible again, says Dr. Amro Al-Saïd Ahmadsaid a computer science professor at Keele University The register. He estimates a total of 38 services were down.
“The incident highlights the dependency effect of modern cloud ecosystems. When a core cloud region experiences outages, the impact extends far beyond the services directly affected. Applications, APIs, databases, authentication systems, payment platforms, analytics pipelines, and third-party SaaS products that depend on these services can all experience cascading outages.”
“An issue in a single cloud region can therefore propagate across the entire digital supply chain, affecting organizations multiple levels away from the original outage. Although this is a regional outage focused on Western Europe2, the scale of services affected highlights a key lesson in cloud engineering: resilience is not achieved simply by moving to the cloud. ®
Gn world
