Kameleoon Main Application Tools Temporarily Unavailable

Post-Mortem Report: Production Infrastructure Service Interruption
Date: July 27, 2026
Impact window: 09:25 AM – 12:00 PM UTC
Total disruption: ~2 hours 35 minutes

Impact
On July 27, between 09:25 AM and 12:00 PM UTC, the Kameleoon application (at app.kameleoon.com) went down and access was not possible. Kameleoon engine files (and SDK config files) were generally available, ensuring correct displaying of AB tests and tracking. However, we had some reports of individual engine files not being served correctly for some customers; in those cases live AB tests were not being delivered until a regeneration process completed.
No customer data was lost or compromised. The deployment was a routine infrastructure configuration update and was not expected to impact customer-facing services.

Cause
The incident occurred during the deployment of a routine production infrastructure configuration update.
During deployment, an incomplete network configuration unintentionally restricted communication between certain infrastructure components. This affected internal platform connectivity and interrupted one of the network paths responsible for serving static assets.
The issue became apparent only after the configuration was applied in the production environment and was not identified during pre-deployment validation.

Resolution
Our engineering teams worked to restore normal platform operation and validate infrastructure health.
Actions taken included:

• Updating the affected network configuration to restore required communication paths
• Restoring connectivity for impacted infrastructure services
• Validating platform health and service recovery across the production environment
• Performing additional verification before returning the platform to normal operation

All services were fully restored by 12:00 PM UTC.
Next Steps & Prevention
To reduce the likelihood of similar incidents in the future, we are implementing the following improvements:

• Strengthened production change governance requiring multiple engineering approvals for production infrastructure changes
• Separation of network configuration deployments from broader cluster configuration updates to reduce deployment scope and risk
• Requiring all production infrastructure changes to follow the standard production release process
• Automated pre-deployment validation to detect unintended configuration differences before rollout