Customer Impact

Between April 17, 2026 and April 19, 2026 at 14:49 UTC, a subset of customers experienced failures when executing serverless jobs in Automation Cloud. Affected customers received “502 Bad Gateway” errors and their serverless jobs failed to execute.

The issue affected serverless workloads across multiple regions. Other Automation Cloud services were not impacted.

Root Cause

The incident was caused by an expired security certificate used for internal service communication in environments supporting serverless workloads. Although a certificate renewal was performed on April 17, a gap in the automated renewal process prevented the updated certificate from being applied to certain serverless environments. As a result, requests routed through the affected gateway were rejected, resulting in the observed “502 Bad Gateway” errors.

Service was restored by rerouting serverless traffic through a healthy gateway while the certificate was renewed across all affected environments.

Detection

The issue was identified on April 19 at 09:16 UTC during the investigation of customer-reported serverless job failures. Engineering traced the failures to the affected gateway and declared an incident, with the response team engaged within minutes.

Response

Following detection on April 19 at 09:16 UTC, engineering began investigating the failed serverless job executions and confirmed that requests routed through the primary gateway were returning “502 Bad Gateway” errors.

The team isolated the affected gateway by validating that job execution was restored when traffic bypassed the gateway, while environments continuing to use the affected gateway still experienced failures. After confirming the fault domain, engineering rerouted serverless traffic to a healthy gateway path, restoring job execution for affected customers. As additional affected environments were identified, the same mitigation was applied until all serverless traffic had been redirected. Service was fully restored by 14:49 UTC, after which the incident was resolved.

Follow-Up

To help prevent similar incidents in the future, we are implementing the following improvements:

• Update the automated certificate renewal process to ensure all environments, including serverless environments, are covered.
• Add proactive certificate health monitoring to detect and alert on certificates approaching expiration.
• Expand monitoring to cover all gateway routing paths to enable faster detection of similar failures.
• Implement staggered certificate expiration schedules across environments so that certificate-related issues are more likely to be identified in lower environments before they can impact production.