Resolved

The Google Cloud incident affecting multiple products in the us-central1 region has been resolved.

According to Google Cloud, the issue was fully resolved for all affected users as of 4 September 2026 at 13:06 PDT (20:06 UTC). Their preliminary root cause analysis indicates that the incident was caused by a faulty change, which was rolled back to mitigate the issue.

We have also re-enabled us-central1 for sandbox provisioning and restored our normal provisioning configuration.

We will work on improving the resilience of our sandbox provisioning process against this type of failure in the future. We apologize for the inconvenience and appreciate your patience throughout the incident.

Update

The issue is still ongoing. Below is the latest update from Google Cloud, published on 4 September 2026 at 18:26:39 UTC+2:

We are investigating an issue with Google Compute Engine, Persistent Disk, Google Cloud Dataflow, Cloud SQL, Google Kubernetes Engine and Cloud Filestore in us-central1 region.

We are experiencing an issue with Google Compute Engine control plane operations (VM creations, deletion, etc) beginning on Friday, 2026-09-04 06:39 PDT. Existing VMs are not affected. Cloud Filestore, Google Kubernetes Engine, Cloud Dataflow are affected.

Our engineering team is attempting a mitigation and continues to investigate the issue.

We will provide an update by Friday, 2026-09-04 10:30 PDT with details.

We continue to track the incident and will share further updates as the situation develops.

Observability data indicates that the issues on the Google Cloud side started escalating again around 15:05 UTC. To prevent continued failures, we have temporarily disabled us-central1 for sandbox provisioning.

During this period, sandboxes for manual plays and hot start pools configured for the us-central1 region will be provisioned in europe-west1 instead.

We apologize for the inconvenience and appreciate your patience while we work to mitigate the impact. We will continue monitoring the situation closely and provide further updates as the status changes.

One of the cloud zones (us-central1-a) where we provision virtual machines is currently experiencing issues. While most machines are still being provisioned within a few seconds, some requests are getting stuck and eventually hitting the timeout threshold.

This primarily affects learners accessing the platform close to the us-central1 region, and the scope is hot start pools referenced to a track or sandbox preset with a VM.

The issue started at 13:17 UTC. It subsided around 13:58 UTC but returned at 14:15 UTC, so the impact has been intermittent rather than a continuous outage.

We apologize for any inconvenience and appreciate your patience as we continue working with Google Cloud Support to resolve the issue.