Axway Status

Unbounded queries overwhelmed the analytics systems in use by Engage and Platform, causing delays in processing between 2PM-3PM UTC today. We will be working to improve query size and performance as well as introduce further restrictions to prevent this from causing issues in the future.

We’re sorry — between 13:00 UTC and 14:00 UTC today some customers experienced partial interruptions to Axway Platform, Axway Analytics, Amplify Marketplace, and Amplify Engage. During that window an internal service error caused elevated failures and automated retries, which temporarily prevented normal access for some users. We restored access by allowing queued requests to proceed while we validated and deployed a tested fix; normal service was confirmed at 14:00 UTC. No customer data was exposed. We’re taking steps to reduce the chance of this happening again, including additional safeguards, improved test coverage, and broader alerting so production issues are surfaced to our public status page and support teams more quickly. If you’re still seeing problems or need help, please contact support through your usual channels.

Description: During the detected outage on Amplify Studio, users experienced instability in activating integrations. Additionally, the receipt of monitoring data from Fusion was affected. Root Cause: The introduction of additional monitoring checks in the environments led to a permissions error on the clusters. This error disrupted connectivity between the Studio and Fusion instances. The standard validation routines prior to being deployed in production didn’t catch this permission mismatch. Corrective Action: The permissions were promptly reverted, and impacted services were verified. To prevent recurrence, we have implemented additional checks to ensure such changes do not affect production without being detected in lower environments first.

This incident has been resolved.

Date: Start Friday 21st - 08:02 AM End Friday 21st - 12:23 PM Impact: The Shared Data Plane experienced an outage, resulting in the interruption of transaction processing. Details: During the startup of new worker nodes two issues were discovered. The startup of the CNI plugin failed and was therefore not initialized. A CNI plugin is required to implement the Kubernetes network model. The second issue was due to the EFS plugin crashing with OOM. EFS is used for persistent storage. To resolve the issue, the memory was increased and unstable nodes were restarted. After these actions the Shared Data Plane returned to a stable state and the outage ended.

Date: February 5-6, 2024 (Start time after 22:00 UTC on February 5th; end time 09:20 UTC on February 6th) Impact: The Shared Data Plane experienced an outage, resulting in the interruption of transaction processing. Description: A thorough investigation of system logs revealed a disruption in network connectivity between nodes within the Hazelcast cluster. Hazelcast serves as the messaging layer between various components in our Orchestrator. The network disruption prevented communication between these critical components, halting message processing and consequently, transaction processing. Corrective Actions & Future Plans: A root cause analysis is underway to determine the specific cause of the network disruption within the Hazelcast cluster. Aside from this analysis, the team is planning to replace Hazelcast with an alternative messaging platform. Enhancements to the monitoring framework are implemented to proactively detect and alert on similar connectivity issues, enabling prompt intervention and minimizing service disruption.

The maintenance has been successfully executed and the EU SaaS Control Plane and EU SaaS Data Plane have been successfully upgraded to the newest version 1.9.0. We appreciate your understanding and patience as we work to enhance our services. If you have any questions or concerns, please do not hesitate to contact our support team.

During a certificate upgrade, a rogue certificate caused a tremendous increase of the CPU of Axway Platform and other services that were inaccessible.

This incident has been resolved.

This incident has been resolved.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.

The scheduled maintenance has been completed.