Summary

On May 1, 2026, customers using the Gmail integration experienced a service event that caused Gmail connections to disappear and the email channel to stop functioning.

The issue affected customers across different regions, more specifically:

• EU-based clients, from 2:40 AM ET.
• US-based clients, from 5 AM ET.

The issue was fully resolved by 9:20 AM ET. No customer data was lost during this event. Messages that were delayed during the incident were fully recovered once service was restored. Current system health is stable.

Root cause

A configuration issue in the infrastructure used by the Gmail integration prevented replacement service tasks from starting correctly during routine task rotation. As running capacity declined over several hours, the Gmail integration became unavailable.

Timeline

Time (ET) Event May 1, ~12:30 AM Service task replacement failures began in production. May 1, ~2:40 AM Engineering was alerted after service capacity dropped to zero in one production environment. May 1, 2:40–9:20 AM Teams investigated the failure, identified the configuration problem, prepared a fix, and deployed it. May 1, 9:20 AM Fix deployed and service restored.

Lessons and improvements

• We are auditing related infrastructure configurations to identify and correct similar patterns in other services.
• We are standardizing how service roles are managed so this class of configuration issue is less likely to recur.
• We are improving alerting so teams are notified earlier when running service capacity drops below expected levels, before a full service interruption occurs.
• We are adding additional checks to catch configuration drift and invalid service role references earlier in the deployment lifecycle.