Google has issued a report and has also apologised for an outage last week that impacted services such as YouTube, Gmail, Cloud, Google, Google Meet, and Google Nest. The outage lasted for a couple of hours, and had also impacted Cloudflare. Google, in its incident summary report, stated that Google Cloud, Google Workspace and Google Security Operations products experienced increased 503 errors in external API requests, impacting customers.
Also read: Google Cloud Outage Impacts Several Users, Cloudflare Acknowledges Issue
“We deeply apologize for the impact this outage has had. Customers and their users trust their businesses to Google, and we will do better. We apologize for the impact this has had not only on our customers’ businesses and their users but also on the trust of our systems. We are committed to making improvements to help avoid outages like this moving forward,” said the Alphabet company on its page.
How the Google Cloud Outage Occurred?
The company has added in its incident response page that Google’s Service Control, the core policy-checking system for Google and Google Cloud APIs, experienced a notable issue following the rollout of a new feature designed for enhanced quota policy checks. Service Control operates regionally, relying on a datastore with metadata that replicates globally to enforce quota policies for Google Cloud and its users.
Also read: Google Cloud Acquires Wiz to Strengthen Multi-cloud Security
On 29 May 2025, Google introduced the new feature, which underwent a region-by-region rollout. However, the specific code path tied to the update was not exercised during the deployment phase, as it required a triggering policy change that had not yet occurred. Although the update included a safeguard to disable the feature if needed, the absence of error handling and feature flag protection exposed a critical vulnerability.
The flaw, specifically a null pointer error, caused the Service Control binary to crash. Feature flags, which typically enable controlled feature rollouts and testing on internal projects, were not implemented for this change. If feature flag protection had been in place, the issue might have been identified and addressed during the staging phase.
Thomas Kurian, CEO, Google Cloud acknowledged the issue stating: “We have been hard at work on the outage today and we are now fully restored across all regions and products. We regret the disruption this caused our customers.”
