Application Performance Degradation

Incident Summary
On June 8–9, 2026, some users experienced errors or were unable to access parts of the application for a period of approximately 18 hours. Because the affected component supports billing and subscription information used across the product, the disruption affected general application access, not only billing pages. No customer data was lost.

 Detection
The issue was identified by internal engineering teams at approximately 4:18 PM ET on June 8, following a significant rise in application errors and reports from Customer Support. While the symptoms were detected within an hour of the initial impact, the underlying cause required extensive investigation as it involved a silent failure of an internal component.

 Root Cause
An internal issue caused the application to lose access to cached billing and subscription data, which led it to generate an unusually high volume of requests to a third-party service we rely on. That volume exceeded the service's allowed limits, and requests were temporarily blocked, producing widespread errors.

 Resolution
Our engineering team corrected the underlying issue, significantly reduced the volume of requests to the third-party service, and restored normal operation. Service recovered by 9:30 AM ET on June 9, and we verified the application was operating normally.

 Preventative Measures
To prevent recurrence, we are implementing the following improvements:

• Preventing the internal cache from disabling itself on transient errors, and alerting immediately if it does
• Reusing authenticated sessions and applying limits earlier in the request path to avoid generating excessive third-party traffic
• Maintaining safe headroom under third-party limits and adding monitoring so this condition is detected within minutes