Google Cloud outage history

A networking issue in Google Cloud's us-central1-b region began around 2:41 PM UTC and led to degraded performance across a broad set of services. Affected products included Google BigQuery, Google Cloud Dataflow, Google Cloud SQL, Cloud Run, App Engine, AlloyDB, Looker, Apigee, Cloud Filestore, and the underlying Hybrid Connectivity and VPC components. The problem was localized to part of the us-central1-b zone, with other zones in the region remaining healthy. Users in affected areas reported intermittent connectivity and elevated packet loss as the issue unfolded.

A regional outage in the Google Cloud us-west1 region caused timeouts, degraded performance, errors, and elevated latencies across a broad set of products. Affected services included Google Cloud BigQuery, Cloud Run, Pub/Sub, Cloud Storage, Cloud SQL, Compute Engine, Google Kubernetes Engine, App Engine, Dataproc, Cloud Data Fusion, Cloud Filestore, AlloyDB for PostgreSQL, Cloud Build, Identity and Access Management, and Persistent Disk, among others. The disruption began around 3:40 PM UTC (08:40 PDT) and impacted customers in the us-west1 region, with user experiences ranging from failed requests to latency spikes.

Impact began around 3:35 AM UTC when Google Cloud reported temperature alerts in europe-west4-a affecting Google Cloud VMware Engine (GCVE), Google Cloud NetApp Volumes, and Bare Metal Solutions (BMS). A datacenter in that region suffered a power failure, triggering a cooling failure. To protect customer data from risks of operating in a high temperature environment, Google proactively reduced workloads while the datacenter operators worked to restore cooling and power and restart affected workloads. The disruption persisted as cooling and power issues were being addressed across the region.

GCVE-stretched cluster customers experienced zonal outages that disrupted network connectivity across multiple regions. GCVE VMs continued to run, but connectivity to those VMs could be unreliable for affected workloads; Storage and Compute services were reported as unaffected. The disruption began 5:00 PM UTC and affected users across regions as engineers worked to identify impacted customers and mitigate the issue.

Investigations identified inter-site communication failures between cluster zones, with a recent configuration update cited as a likely cause. A remediation rollout is in progress, and the connectivity issues persisted during mitigation efforts. The affected VMWare Engine component transitioned from warn to up by 5:35 AM UTC as recovery progressed.

An outage affected Google App Engine Console, blocking access to the Cloud Console and hindering deployment workflows. Users reported being unable to deploy new App Engine versions and access deployment dashboards, with UI issues extending to the Cloud Console. The disruption appeared global, with reports from Europe, India, and the Americas indicating widespread impact on the App Engine Console and related deployment tools.

StatusGator issued an Early Warning Signal beginning at Feb 20, 2026 6:53 AM UTC and ending at Feb 20, 2026 11:23 AM UTC, indicating issues reported by users before the provider acknowledged the problem. There were no official status page updates shown in the data. The early warning window lasted roughly 4 hours and 30 minutes, suggesting a multi-hour impact on affected components.

The outage impacted multiple App Engine components, including the Console and deployment dashboards, with reports continuing through late morning UTC and indicating a broad geographic footprint across several regions.

Google Cloud BigQuery experienced an outage that disrupted querying and data access for users around the world. The BigQuery API and related compute components were affected, causing errors and unavailable results for teams attempting to run queries or access dashboards. The outage lasted from Oct 1, 2025 4:02 PM UTC to Oct 1, 2025 4:36 PM UTC, indicating roughly 34 minutes of degraded to unavailable service. StatusGator observed widespread user-reported issues across multiple regions, signaling a global impact.

An outage affected Google Cloud Gemini API and Vertex AI services, spanning from Sep 29, 2025 6:05 PM UTC to Sep 29, 2025 7:07 PM UTC. Users reported failures when invoking Gemini API endpoints and running Vertex AI workloads, disrupting AI workflows, model deployments, and API integrations across multiple regions. StatusGator detected the issue at Sep 29, 2025 6:05 PM UTC, ahead of any official Google Cloud status-page acknowledgement.

Impact appeared global, with reports from North America, Europe, and Asia indicating widespread disruption to Gemini API and Vertex AI. No official Google Cloud status updates were posted on the status page during the incident in the provided data, and the outage resolved around Sep 29, 2025 7:07 PM UTC.

Google Cloud reported an incident in the us-east1 region beginning at Jul 18, 2025 3:55 PM UTC, with elevated latency and error rates across a broad set of Google Cloud services. The disruption affected dozens of services including Google Compute Engine, Cloud Run, Cloud Storage, Cloud SQL, Google Kubernetes Engine, BigQuery, Dataflow, Pub/Sub, Identity and Access Management, Secret Manager, and more, resulting in degraded performance and intermittent outages for workloads in us-east1. The outage lasted a little over two hours, with recovery progress noted as engineers addressed a hardware infrastructure failure in the region and began restoring the network control plane and service availability.

Cloud Build experienced issues where builds were failing to start or remaining queued, disrupting CI/CD workflows for affected users. StatusGator detected the issue at Jul 10, 2025 7:07 PM UTC with the early warning signal ending at 7:26 PM UTC. User reports started around 7:00 PM UTC from multiple regions, including Columbus, Los Angeles, La Quinta, Wilbraham, and Zagreb, Croatia, indicating broad geographic impact.

There were no official status updates posted on Google Cloud's Status Page during this window. The incident is classified as minor, reflecting degraded build/startup performance rather than a full outage, with builds either failing to start or stalling in the queue and causing delays across affected projects.

Google Cloud's Locker Studio experienced an outage with users across Asia-Pacific and other regions reporting the service not loading and Looker Studio files failing to load. The issue produced slow performance and access problems for dashboards and reports, affecting a wide range of users in cities including Jakarta, Hanoi, Ho Chi Minh City, Hong Kong, Sydney, Melbourne, and Auckland.

StatusGator logged an Early Warning Signal starting at Jun 24, 2025 2:10 AM UTC as user reports mounted, with the event lasting about 2 hours and 35 minutes until Jun 24, 2025 4:45 AM UTC. There were no official status updates published by Google Cloud for this incident.