Proton, the privacy-focused tech company, experienced a significant worldwide outage today that impacted its suite of services, including Proton Mail, Proton Calendar, ProtonVPN, Proton Drive, Proton Pass, and Proton Wallet.
The outage, which began around 10:00 AM ET, left many users unable to access their accounts due to unexpected network issues.
Timeline of the Outage
- 10:00 AM ET: Proton services began experiencing connectivity issues. The company acknowledged the problem, stating on its status page: “We are currently experiencing intermittent network issues affecting some of our users. We are working to fully restore services as soon as possible.”
- 12:37 PM ET: Proton Mail connectivity was restored, bringing back a crucial service for many users.
- 1:27 PM ET: Proton announced that all services, including Proton Calendar, were fully operational.
Users affected by the outage reported seeing error messages like:
“Something went wrong. We couldn’t load this page. Please refresh the page or check your internet connection.”
Root Cause of the Issue
In an incident report, Proton revealed that the outage was triggered by a sudden surge in database server connections globally. The spike overwhelmed the company’s infrastructure, which is currently undergoing a transition to a Kubernetes-based system.
While ProtonVPN, Proton Pass, and other services were quickly restored, Proton Mail and Proton Calendar experienced prolonged issues. Proton explained that the ongoing migration limited their ability to dynamically scale capacity during the disruption.
A software change was identified as a likely factor in the load spike. Though it had been implemented earlier, its effects only became apparent today. Rolling back the change resolved the issue, and the database load returned to normal.
Resolved
The outage took approximately two hours to fully resolve, with degraded performance improving progressively during the recovery window. Proton has assured users that completing the Kubernetes migration will enhance the infrastructure’s resilience and restore higher levels of redundancy, reducing the risk of similar incidents.
