Intermittent API errors across US

Update timeline

1. identified Mar 24, 2026, 10:35 AM UTC

   We have identified an issue in our network configuration that is leading to some requests failing. We've identified fix and are deploying to production.

2. identified Mar 24, 2026, 10:35 AM UTC

   We are continuing to work on a fix for this issue.

3. identified Mar 24, 2026, 10:36 AM UTC

   We are continuing to work on a fix for this issue.

4. monitoring Mar 24, 2026, 11:34 AM UTC

   We have deployed a fix to our network configuration and all services are healthy. We will continue monitoring.

5. resolved Mar 24, 2026, 12:19 PM UTC

   We've seen no errors related to network configuration.

6. postmortem Mar 24, 2026, 12:21 PM UTC

   **Root cause** Our cloud networking layer had insufficient capacity to handle the volume of outbound connections from our services. As traffic grew, we periodically exceeded the available connection limits, causing some requests to be dropped. **What we fixed** We restructured our internal network architecture, significantly reducing the load on our outbound connection layer. This freed up capacity. ‌ **Ensuring it won’t happen again** We’ve permanently re-architected to prevent this specific issue. We’ve added monitoring \+ alerts to catch similar network errors.